servalcat 0.4.131__cp314-cp314t-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl

servalcat/utils/commands.py

@@ -0,0 +1,1629 @@
+"""
+Author: "Keitaro Yamashita, Garib N. Murshudov"
+MRC Laboratory of Molecular Biology
+    
+This software is released under the
+Mozilla Public License, version 2.0; see LICENSE.
+"""
+from __future__ import absolute_import, division, print_function, generators
+from servalcat.utils import logger
+from servalcat.utils import fileio
+from servalcat.utils import symmetry
+from servalcat.utils import model
+from servalcat.utils import hkl
+from servalcat.utils import restraints
+from servalcat.utils import maps
+from servalcat.refmac import refmac_keywords
+from servalcat.refine.refine import Geom, RefineParams
+from servalcat import ext
+import os
+import gemmi
+import numpy
+import scipy.spatial
+import pandas
+import json
+import re
+import argparse
+
+def add_arguments(p):
+    subparsers = p.add_subparsers(dest="subcommand")
+
+    # show
+    parser = subparsers.add_parser("show", description = 'Show file info supported by the program')
+    parser.add_argument('files', nargs='+')
+
+    # json2csv
+    parser = subparsers.add_parser("json2csv", description = 'Convert json to csv for plotting')
+    parser.add_argument('json')
+    parser.add_argument('-o', '--output_prefix')
+
+    # symmodel
+    parser = subparsers.add_parser("symmodel", description="Add symmetry annotation to model")
+    parser.add_argument('--model', required=True)
+    group = parser.add_mutually_exclusive_group()
+    group.add_argument('--map', help="Take box size from the map")
+    group.add_argument('--cell', type=float, nargs=6, metavar=("a", "b", "c", "alpha", "beta", "gamma"),
+                       help="Box size")
+    sym_group = parser.add_argument_group("symmetry")
+    symmetry.add_symmetry_args(sym_group, require_pg=True)
+    parser.add_argument('--contacting_only', action="store_true", help="Filter out non-contacting NCS")
+    parser.add_argument('--chains', nargs="*", action="append", help="Select chains to keep")
+    parser.add_argument('--howtoname', choices=["dup", "short", "number"], default="short",
+                        help="How to decide new chain IDs in expanded model (default: short); "
+                        "dup: use original chain IDs (with different segment IDs), "
+                        "short: use unique new IDs, "
+                        "number: add number to original chain ID")
+    parser.add_argument('--biomt', action="store_true", help="Add BIOMT also")
+    parser.add_argument('-o', '--output_prfix')
+    parser.add_argument('--pdb', action="store_true", help="Write a pdb file")
+    parser.add_argument('--cif', action="store_true", help="Write a cif file")
+
+    # helical_biomt
+    parser = subparsers.add_parser("helical_biomt", description="generate BIOMT of helical reconstruction for PDB deposition")
+    parser.add_argument('--model', required=True)
+    group = parser.add_mutually_exclusive_group()
+    group.add_argument('--map', help="Take box size from the map")
+    group.add_argument('--cell', type=float, nargs=6, metavar=("a", "b", "c", "alpha", "beta", "gamma"),
+                       help="Box size")
+    sym_group = parser.add_argument_group("symmetry")
+    symmetry.add_symmetry_args(sym_group, require_pg=True)
+    parser.add_argument('--start', type=int)
+    parser.add_argument('--end', type=int)
+    parser.add_argument('--howtoname', choices=["dup", "short", "number"], default="short",
+                        help="How to decide new chain IDs in expanded model (default: short); "
+                        "dup: use original chain IDs (with different segment IDs), "
+                        "short: use unique new IDs, "
+                        "number: add number to original chain ID")
+    parser.add_argument('-o', '--output_prfix')
+
+    # expand
+    parser = subparsers.add_parser("expand", description="Expand symmetry")
+    parser.add_argument('--model', required=True)
+    parser.add_argument('--chains', nargs="*", action="append", help="Select chains to keep")
+    group = parser.add_mutually_exclusive_group()
+    group.add_argument('--howtoname', choices=["dup", "short", "number"], default="short",
+                       help="How to decide new chain IDs in expanded model (default: short); "
+                       "dup: use original chain IDs (with different segment IDs), "
+                       "short: use unique new IDs, "
+                       "number: add number to original chain ID")
+    group.add_argument("--split", action="store_true", help="split file for each operator")
+    parser.add_argument('-o', '--output_prfix')
+    parser.add_argument('--pdb', action="store_true", help="Write a pdb file")
+    parser.add_argument('--cif', action="store_true", help="Write a cif file")
+
+    # h_add
+    parser = subparsers.add_parser("h_add", description = 'Add hydrogen in riding position')
+    parser.add_argument('model')
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument('-o','--output')
+    parser.add_argument("--pos", choices=["elec", "nucl"], default="elec")
+
+    # add_op3
+    parser = subparsers.add_parser("add_op3", description = "Add OP3 atoms to 5' ends")
+    parser.add_argument('model')
+    parser.add_argument('--chains', nargs="*", action="append", help="For selected chains only")
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument('-o','--output')
+
+    # map_peaks
+    parser = subparsers.add_parser("map_peaks", description = 'List density peaks and write a coot script')
+    parser.add_argument('--model', required=True, help="Model")
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--map', help="Map file")
+    group.add_argument('--mtz', help="MTZ for map file")
+    parser.add_argument('--mtz_labels', default="DELFWT,PHDELWT", help='F,PHI labels (default: %(default)s)')
+    parser.add_argument('--oversample_pixel', type=float, help='Desired pixel spacing in map (Angstrom)')
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--sigma_level', type=float, help="Threshold map level in sigma unit")
+    group.add_argument('--abs_level', type=float, help="Threshold map level in absolute unit")
+    parser.add_argument('--blob_pos', choices=["peak", "centroid"], default="centroid",
+                       help="default: %(default)s")
+    parser.add_argument('--min_volume', type=float, default=0.3, help="minimum blob volume (default: %(default).1f)")
+    parser.add_argument('--max_volume', type=float, help="maximum blob volume (default: none)")
+    parser.add_argument('-o','--output_prefix', default="peaks")
+    
+    # h_density
+    parser = subparsers.add_parser("h_density", description = 'Hydrogen density analysis')
+    parser.add_argument('--model', required=True, help="Model with hydrogen atoms")
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--map', help="Fo-Fc map file")
+    group.add_argument('--mtz', help="MTZ for Fo-Fc map file")
+    parser.add_argument('--mtz_labels', default="DELFWT,PHDELWT", help='F,PHI labels (default: %(default)s)')
+    parser.add_argument('--oversample_pixel', type=float, help='Desired pixel spacing in map (Angstrom)')
+    #parser.add_argument("--source", choices=["electron", "xray", "neutron"], default="electron")
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--sigma_level', type=float, help="Threshold map level in sigma unit")
+    group.add_argument('--abs_level', type=float, help="Threshold map level in absolute unit")
+    parser.add_argument('--max_dist', type=float, default=0.5, help="max distance between peak and hydrogen position in the model (default: %(default).1f)")
+    parser.add_argument('--blob_pos', choices=["peak", "centroid"], default="centroid",
+                       help="default: %(default)s")
+    parser.add_argument('--min_volume', type=float, default=0.3, help="minimum blob volume (default: %(default).1f)")
+    parser.add_argument('--max_volume', type=float, default=3, help="maximum blob volume (default: %(default).1f)")
+    parser.add_argument('-o','--output_prefix')
+
+    # fix_link
+    parser = subparsers.add_parser("fix_link", description = 'Fix LINKR/_struct_conn records in the model')
+    parser.add_argument('model')
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument('--bond_margin', type=float, default=1.3, help='(default: %(default).1f)')
+    parser.add_argument('--metal_margin', type=float, default=1.1, help='(default: %(default).1f)')
+    parser.add_argument('-o','--output', help="Default: input_fixlink.{pdb|mmcif}")
+
+    # merge_models
+    parser = subparsers.add_parser("merge_models", description = 'Merge multiple model files')
+    parser.add_argument('models', nargs="+")
+    parser.add_argument('-o','--output', required=True)
+
+    # merge_dicts
+    parser = subparsers.add_parser("merge_dicts", description = 'Merge restraint dictionary cif files')
+    parser.add_argument('cifs', nargs="+")
+    parser.add_argument('-o','--output', default="merged.cif", help="Output cif file (default: %(default)s)")
+
+    # geom
+    parser = subparsers.add_parser("geom", description = 'Calculate geometry and show outliers')
+    parser.add_argument('model')
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument('--keywords', nargs='+', action="append",
+                        help="refmac keyword(s)")
+    parser.add_argument('--keyword_file', nargs='+', action="append",
+                        help="refmac keyword file(s)")
+    parser.add_argument('--sigma', type=float, default=5,
+                        help="sigma cutoff to print outliers (default: %(default).1f)")
+    parser.add_argument('--per_atom_score_as_b', action='store_true',
+                        help="write model file with per-atom score as B values")
+    parser.add_argument("--check_skew", action='store_true', help="(experimental) check bond skew to test magnification")
+    parser.add_argument('-n', '--nucleus', action="store_true", help="Use nucleus distances (for neutron)")
+    parser.add_argument("--ignore_h", action='store_true', help="ignore hydrogen")
+    parser.add_argument("--selection", help="evaluate part of the model")
+    parser.add_argument('--dump_all', action="store_true", help=argparse.SUPPRESS)
+    parser.add_argument('-o', '--output_prefix', 
+                        help="default: taken from input file")
+
+    # conf
+    parser = subparsers.add_parser("conf", description = 'Compare conformations')
+    parser.add_argument('models', nargs="+")
+    parser.add_argument("--min_diff", type=float, default=60.)
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument("--same_chain", action='store_true', help="Only between same chains (more than one file)")
+    parser.add_argument('-o', '--output_prefix', default="conf",
+                        help="")
+
+    # adp
+    parser = subparsers.add_parser("adp", description = 'ADP analysis')
+    parser.add_argument('model')
+    parser.add_argument('-o', '--output_prefix',
+                        help="default: taken from input file")
+    
+    # power
+    parser = subparsers.add_parser("power", description = 'Show power spectrum')
+    parser.add_argument("--map",  nargs="*", action="append")
+    parser.add_argument("--halfmaps",  nargs="*", action="append")
+    parser.add_argument('--mask', help='Mask file')
+    parser.add_argument('-d', '--resolution', type=float)
+    parser.add_argument('-o', '--output_prefix', default="power")
+
+    # fcalc
+    parser = subparsers.add_parser("fcalc", description = 'Structure factor from model')
+    parser.add_argument('--model', required=True)
+    parser.add_argument("--no_expand_ncs", action='store_true', help="Do not expand strict NCS in MTRIX or _struct_ncs_oper")
+    parser.add_argument("--method", choices=["fft", "direct"], default="fft")
+    parser.add_argument("--source", choices=["electron", "xray", "neutron", "custom"], default="electron")
+    parser.add_argument('--ligand', nargs="*", action="append")
+    parser.add_argument("--monlib",
+                        help="Monomer library path. Default: $CLIBD_MON")
+    parser.add_argument('--cell', type=float, nargs=6, metavar=("a", "b", "c", "alpha", "beta", "gamma"),
+                        help="Override unit cell")
+    parser.add_argument('--auto_box_with_padding', type=float, help="Determine box size from model with specified padding")
+    parser.add_argument('--cutoff', type=float, default=1e-5)
+    parser.add_argument('--rate', type=float, default=1.5)
+    parser.add_argument('--add_dummy_sigma', action='store_true', help="write dummy SIGF")
+    parser.add_argument('--as_intensity', action='store_true', help="if you want |F|^2")
+    parser.add_argument('--keep_charges',  action='store_true',
+                        help="Use scattering factor for charged atoms. Use it with care.")
+    parser.add_argument('-d', '--resolution', type=float, required=True)
+    parser.add_argument('-o', '--output_prefix')
+
+    # nemap
+    parser = subparsers.add_parser("nemap", description = 'Normalized expected map calculation from half maps')
+    parser.add_argument("--halfmaps", required=True, nargs=2)
+    parser.add_argument('--pixel_size', type=float, help='Override pixel size (A)')
+    parser.add_argument("--half1_only", action='store_true', help="Only use half 1 for map calculation (use half 2 only for noise estimation)")
+    parser.add_argument('-B', type=float, help="local B value")
+    parser.add_argument("--no_fsc_weights", action='store_true',
+                        help="Just for debugging purpose: turn off FSC-based weighting")
+    parser.add_argument("--sharpening_b", type=float,
+                        help="Use B value (negative value for sharpening) instead of standard deviation of the signal")
+    parser.add_argument("-d", '--resolution', type=float)
+    parser.add_argument('-m', '--mask', help="mask file")
+    parser.add_argument('-o', '--output_prefix', default='nemap')
+    parser.add_argument("--trim", action='store_true', help="Write trimmed maps")
+    parser.add_argument("--trim_mtz", action='store_true', help="Write trimmed mtz")
+    parser.add_argument("--local_fourier_weighting_with", type=float, default=0,
+                        help="Experimental: give kernel size in A^-1 unit to use local Fourier weighting instead of resolution-dependent weights")
+
+    # blur
+    parser = subparsers.add_parser("blur", description = 'Blur data by specified B value')
+    parser.add_argument('--hklin', required=True, help="input MTZ file")
+    parser.add_argument('-B', type=float, required=True, help="B value for blurring (negative value for sharpening)")
+    parser.add_argument('-o', '--output_prefix')
+
+    # mask_from_model
+    parser = subparsers.add_parser("mask_from_model", description = 'Make a mask from model')
+    parser.add_argument("--map", required=True, help="For unit cell and pixel size reference")
+    parser.add_argument("--model", required=True)
+    parser.add_argument("--selection")
+    parser.add_argument('--radius', type=float, required=True,
+                        help='Radius in angstrom')
+    parser.add_argument('--soft_edge', type=float, default=0,
+                        help='Soft edge (default: %(default).1f)')
+    parser.add_argument('-o', '--output', default="mask_from_model.mrc")
+    
+    # applymask (and normalize within mask)
+    parser = subparsers.add_parser("applymask", description = 'Apply mask and optionally normalize map within mask')
+    parser.add_argument("--map", required=True)
+    parser.add_argument('--mask', required=True, help='Mask file')
+    parser.add_argument("--normalize", action='store_true',
+                        help="Normalize map values using mean and sd within the mask")
+    parser.add_argument("--trim", action='store_true', help="Write trimmed map")
+    parser.add_argument('--mask_cutoff', type=float, default=0.5,
+                        help="cutoff value for normalization and trimming (default: %(default)s)")
+    parser.add_argument('-o', '--output_prefix')
+
+    # map2mtz
+    parser = subparsers.add_parser("map2mtz", description = 'FFT map and write an mtz')
+    parser.add_argument("--map", required=True)
+    parser.add_argument("-d", '--resolution', type=float)
+    parser.add_argument('-o', '--output')
+
+    # sm2mm
+    parser = subparsers.add_parser("sm2mm", description = 'Small molecule files (cif/hkl/res/ins) to macromolecules (pdb/mmcif/mtz)')
+    parser.add_argument('files', nargs='+', help='Cif/ins/res/hkl files')
+    parser.add_argument('-o', '--output_prefix')
+
+    # mm2ins
+    parser = subparsers.add_parser("mm2ins", description = 'convert pdb/mmcif to ins for shelxl/olex2')
+    parser.add_argument('model')
+    parser.add_argument('--hklin')
+    parser.add_argument('-o', '--output')
+
+    # seq
+    parser = subparsers.add_parser("seq", description = 'Print/align model sequence')
+    parser.add_argument("--model", required=True)
+    parser.add_argument('--seq', nargs="*", action="append", help="Sequence file(s)")
+    parser.add_argument('--scoring', nargs=6, type=int, default=(1, 0, -1, -1, 0, -1),
+                        metavar=("match", "mismatch", "gapo", "gape", "good_gapo", "bad_gapo"),
+                        help="scoring function. default: %(default)s")
+
+    # dnarna
+    parser = subparsers.add_parser("dnarna", description = 'DNA to RNA or RNA to DNA model conversion')
+    parser.add_argument("model")
+    group = parser.add_mutually_exclusive_group(required=True)
+    group.add_argument('--to_dna', action='store_true', help="To DNA")
+    group.add_argument('--to_rna', action='store_true', help="To RNA")
+    parser.add_argument('--chains', nargs="*", action="append", help="Select chains to convert")
+    parser.add_argument('-o', '--output')
+
+# add_arguments()
+
+def parse_args(arg_list):
+    parser = argparse.ArgumentParser()
+    add_arguments(parser)
+    return parser.parse_args(arg_list)
+# parse_args()
+
+def symmodel(args):
+    if args.chains: args.chains = sum(args.chains, [])
+    model_format = fileio.check_model_format(args.model)
+
+    howtoname = dict(dup=gemmi.HowToNameCopiedChain.Dup,
+                     short=gemmi.HowToNameCopiedChain.Short,
+                     number=gemmi.HowToNameCopiedChain.AddNumber)[args.howtoname]
+
+    if (args.twist, args.rise).count(None) == 1:
+        raise SystemExit("ERROR: give both helical parameters --twist and --rise")
+
+    is_helical = args.twist is not None
+    st, cif_ref = fileio.read_structure_from_pdb_and_mmcif(args.model)
+    st.spacegroup_hm = "P 1"
+    map_and_start = None
+    if args.map:
+        logger.writeln("Reading cell from map")
+        map_and_start = fileio.read_ccp4_map(args.map, header_only=True)
+        st.cell = map_and_start[0].unit_cell
+    elif args.cell:
+        st.cell = gemmi.UnitCell(*args.cell)
+    elif not st.cell.is_crystal():
+        raise SystemExit("Error: Unit cell parameters look wrong. Please use --map or --cell")
+
+    if args.chains:
+        logger.writeln("Keep {} chains only".format(" ".join(args.chains)))
+        chains = set(args.chains)
+        for m in st:
+            to_del = [c.name for c in m if c.name not in chains]
+            for c in to_del: m.remove_chain(c)
+        if st[0].count_atom_sites() == 0:
+            raise SystemExit("ERROR: no atoms left. Check --chains option.")
+
+    all_chains = [c.name for c in st[0] if c.name not in st[0]]
+
+    symmetry.update_ncs_from_args(args, st, map_and_start=map_and_start, filter_contacting=args.contacting_only)
+
+    if args.biomt:
+        st.assemblies.clear()
+        st.raw_remarks = []
+        a = model.prepare_assembly("1", all_chains, st.ncs, is_helical=is_helical)
+        st.assemblies.append(a)
+
+    if not args.output_prfix:
+        args.output_prfix = fileio.splitext(os.path.basename(args.model))[0] + "_asu"
+
+    if args.pdb or args.cif:
+        fileio.write_model(st, args.output_prfix, pdb=args.pdb, cif=args.cif, cif_ref=cif_ref)
+    else:
+        fileio.write_model(st, file_name=args.output_prfix+model_format, cif_ref=cif_ref)
+
+    # Sym expand
+    model.expand_ncs(st, howtoname=howtoname)
+    st.assemblies.clear()
+    args.output_prfix += "_expanded"
+    if args.pdb or args.cif:
+        fileio.write_model(st, args.output_prfix, pdb=args.pdb, cif=args.cif)
+    else:
+        fileio.write_model(st, file_name=args.output_prfix+model_format)
+# symmodel()
+
+def helical_biomt(args):
+    if (args.twist, args.rise).count(None) > 0:
+        raise SystemExit("ERROR: give helical parameters --twist and --rise")
+
+    model_format = fileio.check_model_format(args.model)
+    howtoname = dict(dup=gemmi.HowToNameCopiedChain.Dup,
+                     short=gemmi.HowToNameCopiedChain.Short,
+                     number=gemmi.HowToNameCopiedChain.AddNumber)[args.howtoname]
+
+    st, cif_ref = fileio.read_structure_from_pdb_and_mmcif(args.model)
+    st.spacegroup_hm = "P 1"
+    map_and_start = None
+    if args.map:
+        logger.writeln("Reading cell from map")
+        map_and_start = fileio.read_ccp4_map(args.map, header_only=True)
+        st.cell = map_and_start[0].unit_cell
+    elif args.cell:
+        st.cell = gemmi.UnitCell(*args.cell)
+    elif not st.cell.is_crystal():
+        raise SystemExit("Error: Unit cell parameters look wrong. Please use --map or --cell")
+
+    all_chains = [c.name for c in st[0] if c.name not in st[0]]
+
+    ncsops = symmetry.ncsops_from_args(args, st.cell, map_and_start=map_and_start, st=st,
+                                       helical_min_n=args.start, helical_max_n=args.end)
+    #ncsops = [x for x in ncsops if not x.tr.is_identity()] # remove identity
+
+    logger.writeln("")
+    logger.writeln("-------------------------------------------------------------")
+    logger.writeln("You may need to write following matrices in OneDep interface:")
+    for idx, op in enumerate(ncsops):
+        logger.writeln("")
+        logger.writeln("operator {}".format(idx+1))
+        mat = op.tr.mat.tolist()
+        vec = op.tr.vec.tolist()
+        for i in range(3):
+            mstr = ["{:10.6f}".format(mat[i][j]) for j in range(3)]
+            logger.writeln("{} {:14.5f}".format(" ".join(mstr), vec[i]))
+    logger.writeln("-------------------------------------------------------------")
+    logger.writeln("")
+
+    # BIOMT
+    st.assemblies.clear()
+    st.raw_remarks = []
+    a = model.prepare_assembly("1", all_chains, ncsops, is_helical=True)
+    st.assemblies.append(a)
+
+    if not args.output_prfix:
+        args.output_prfix = fileio.splitext(os.path.basename(args.model))[0] + "_biomt"
+
+    fileio.write_model(st, args.output_prfix, pdb=(model_format == ".pdb"), cif=True, cif_ref=cif_ref)
+    logger.writeln("")
+    logger.writeln("These {}.* files may be used for deposition (once OneDep implemented reading BIOMT from file..)".format(args.output_prfix))
+    logger.writeln("")
+    # BIOMT expand
+    st.transform_to_assembly("1", howtoname)
+    args.output_prfix += "_expanded"
+    fileio.write_model(st, file_name=args.output_prfix+model_format)
+    logger.writeln(" note that this expanded model file is just for visual inspection, *not* for deposition!")
+# helical_biomt()
+
+def symexpand(args):
+    if args.chains: args.chains = sum(args.chains, [])
+    model_format = fileio.check_model_format(args.model)
+    if not args.split:
+        howtoname = dict(dup=gemmi.HowToNameCopiedChain.Dup,
+                         short=gemmi.HowToNameCopiedChain.Short,
+                         number=gemmi.HowToNameCopiedChain.AddNumber)[args.howtoname]
+
+    st = fileio.read_structure(args.model)
+
+    if args.chains:
+        logger.writeln("Keep {} chains only".format(" ".join(args.chains)))
+        chains = set(args.chains)
+        for m in st:
+            to_del = [c.name for c in m if c.name not in chains]
+            for c in to_del: m.remove_chain(c)
+
+    all_chains = [c.name for c in st[0] if c.name not in st[0]]
+
+    if not args.output_prfix:
+        args.output_prfix = fileio.splitext(os.path.basename(args.model))[0]
+
+    if len(st.ncs) > 0:
+        symmetry.show_ncs_operators_axis_angle(st.ncs)
+        non_given = [op for op in st.ncs if not op.given]
+        if len(non_given) > 0:
+            if args.split:
+                for i, op in enumerate(st.ncs):
+                    if op.given: continue
+                    st_tmp = st.clone()
+                    for m in st_tmp: m.transform_pos_and_adp(op.tr)
+                    output_prfix = args.output_prfix + "_ncs_{:02d}".format(i+1)
+                    if args.pdb or args.cif:
+                        fileio.write_model(st_tmp, output_prfix, pdb=args.pdb, cif=args.cif)
+                    else:
+                        fileio.write_model(st_tmp, file_name=output_prfix+model_format)
+            else:
+                st_tmp = st.clone()
+                model.expand_ncs(st_tmp, howtoname=howtoname)
+                output_prfix = args.output_prfix + "_ncs_expanded"
+                if args.pdb or args.cif:
+                    fileio.write_model(st_tmp, output_prfix, pdb=args.pdb, cif=args.cif)
+                else:
+                    fileio.write_model(st_tmp, file_name=output_prfix+model_format)
+        else:
+            logger.writeln("All operators are already expanded (marked as given). Exiting.")
+    else:
+        logger.writeln("No NCS operators found. Exiting.")
+    
+    if len(st.assemblies) > 0: # should we support BIOMT?
+        pass
+# symexpand()
+
+def h_add(args):
+    st = fileio.read_structure(args.model)
+    model_format = fileio.check_model_format(args.model)
+    
+    if not args.output:
+        tmp = fileio.splitext(os.path.basename(args.model))[0]
+        args.output = tmp + "_h" + model_format
+    logger.writeln("Output file: {}".format(args.output))
+        
+    args.ligand = sum(args.ligand, []) if args.ligand else []
+    monlib = restraints.load_monomer_library(st,
+                                             monomer_dir=args.monlib,
+                                             cif_files=args.ligand)
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)
+    restraints.find_and_fix_links(st, monlib, find_metal_links=False, add_found=False)
+    try:
+        restraints.add_hydrogens(st, monlib, args.pos)
+    except RuntimeError as e:
+        raise SystemExit("Error: {}".format(e))
+
+    fileio.write_model(st, file_name=args.output)
+# h_add()
+
+def add_op3(args):
+    if args.chains: args.chains = sum(args.chains, [])
+    st = fileio.read_structure(args.model)
+    model_format = fileio.check_model_format(args.model)
+    
+    if not args.output:
+        tmp = fileio.splitext(os.path.basename(args.model))[0]
+        args.output = tmp + "_op3" + model_format
+    logger.writeln("Output file: {}".format(args.output))
+    
+    args.ligand = sum(args.ligand, []) if args.ligand else []
+    monlib = restraints.load_monomer_library(st,
+                                             monomer_dir=args.monlib,
+                                             cif_files=args.ligand)
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)
+
+    for chain in st[0]:
+        if args.chains and chain.name not in args.chains: continue
+        p = chain.get_polymer()
+        if not p: continue
+        p_type = p.check_polymer_type()
+        if p_type not in (gemmi.PolymerType.Dna, gemmi.PolymerType.Rna): continue
+        r0 = p[0]
+        # TODO: alias
+        # TODO: altlocs
+        alt = "*"
+        if r0.find_atom("OP3", alt): continue
+        a_op1 = r0.find_atom("OP1", alt)
+        a_op2 = r0.find_atom("OP2", alt)
+        a_o5p = r0.find_atom("O5'", alt)
+        a_p = r0.find_atom("P", alt)
+        if None in (a_op1, a_op2, a_o5p, a_p):
+            logger.writeln(f"Error: atoms not found. skipping {chain.name}/{r0}")
+            continue
+        logger.writeln(f"Adding OP3 to {chain.name}/{r0}")
+        a_op3 = r0.add_atom(a_p) # inherit ADP and occupancy
+        a_op3.name = "OP3"
+        a_op3.element = gemmi.Element("O")
+        v1 = a_p.pos - a_op1.pos
+        v2 = a_p.pos - a_op2.pos
+        v3 = a_p.pos - a_o5p.pos
+        v = v1 + v2 + v3
+        a_op3.pos = a_p.pos + v / v.length() * 1.517
+    
+    fileio.write_model(st, file_name=args.output)
+# add_op3()
+
+def read_map_and_oversample(map_in=None, mtz_in=None, mtz_labs=None, oversample_pixel=None):
+    if mtz_in is not None:
+        mtz = fileio.read_mmhkl(mtz_in)
+        lab_f, lab_phi = mtz_labs.split(",")
+        asu = mtz.get_f_phi(lab_f, lab_phi)
+        if oversample_pixel is not None:
+            d_min = numpy.min(asu.make_d_array())
+            sample_rate = d_min / oversample_pixel
+        else:
+            sample_rate = 3
+        gr = asu.transform_f_phi_to_map(sample_rate=sample_rate)
+    elif map_in is not None:
+        gr = fileio.read_ccp4_map(map_in)[0]
+        if oversample_pixel is not None:
+            asu = gemmi.transform_map_to_f_phi(gr).prepare_asu_data()
+            d_min = numpy.min(asu.make_d_array())
+            sample_rate = d_min / oversample_pixel
+            gr = asu.transform_f_phi_to_map(sample_rate=sample_rate)
+    else:
+        raise SystemExit("Invalid input")
+    
+    if oversample_pixel is not None:
+        logger.writeln("--oversample_pixel= {} is requested.".format(oversample_pixel))
+        logger.writeln(" recalculated grid:")
+        logger.writeln("  {:4d} {:4d} {:4d}".format(*gr.shape))
+        logger.writeln(" spacings:")
+        logger.writeln("  {:.6f} {:.6f} {:.6f}".format(*gr.spacing))
+        #maps.write_ccp4_map("{}_oversampled.mrc".format(output_prefix), gr)
+
+    return gr
+# read_map_and_oversample()    
+
+def map_peaks(args):
+    st = fileio.read_structure(args.model)
+    gr = read_map_and_oversample(map_in=args.map, mtz_in=args.mtz, mtz_labs=args.mtz_labels,
+                                 oversample_pixel=args.oversample_pixel)
+    gr_sigma = numpy.std(gr)
+    if args.abs_level is not None:
+        cutoff = args.abs_level
+    else:
+        cutoff = args.sigma_level * gr_sigma # assuming mean(gr) = 0
+        
+    blobs = gemmi.find_blobs_by_flood_fill(gr, cutoff,
+                                           min_volume=args.min_volume, min_score=0)
+    blobs.extend(gemmi.find_blobs_by_flood_fill(gr, cutoff, negate=True,
+                                                min_volume=args.min_volume, min_score=0))
+    getpos = dict(peak=lambda x: x.peak_pos,
+                  centroid=lambda x: x.centroid)[args.blob_pos]
+    st_peaks = model.st_from_positions([getpos(b) for b in blobs])
+    st_peaks.cell = st.cell
+    st_peaks.ncs = st.ncs
+    st_peaks.setup_cell_images()
+    logger.writeln("{} peaks detected".format(len(blobs)))
+    #st_peaks.write_pdb("peaks.pdb")
+    
+    # Filter symmetry related
+    ns = gemmi.NeighborSearch(st_peaks[0], st_peaks.cell, 5.).populate()
+    cs = gemmi.ContactSearch(1.)
+    cs.ignore = gemmi.ContactSearch.Ignore.SameAsu
+    results = cs.find_contacts(ns)
+    del_idxes = set()
+    for r in results:
+        if r.partner1.residue.seqid.num not in del_idxes:
+            del_idxes.add(r.partner2.residue.seqid.num)
+    for i in reversed(sorted(del_idxes)):
+        del st_peaks[0][0][i]
+        del blobs[i]
+    #st_peaks.write_pdb("peaks_asu.pdb")
+    logger.writeln("{} peaks after removing symmetry equivalents".format(len(blobs)))
+
+    # Assign to nearest atom
+    ns = gemmi.NeighborSearch(st[0], st.cell, 10.).populate() # blob is rejected if > 10 A. ok?
+    peaks = []
+    for b in blobs:
+        bpos = getpos(b)
+        map_val = gr.interpolate_value(bpos)
+        if (args.max_volume is not None and b.volume > args.max_volume) or abs(map_val) < cutoff: continue
+        x = ns.find_nearest_atom(bpos)
+        if x is None: # this should not happen
+            logger.writeln("no nearest atom: value={:.2e} volume= {:.2f} pos= {}".format(map_val, b.volume, bpos))
+            continue
+        chain = st[0][x.chain_idx]
+        res = chain[x.residue_idx]
+        atom = res[x.atom_idx]
+        im = st.cell.find_nearest_image(atom.pos, bpos, gemmi.Asu.Any)
+        if st.cell.is_crystal():
+            bpos = st.cell.find_nearest_pbc_position(atom.pos, bpos, im.sym_idx)
+        elif im.sym_idx > 0:
+            bpos = st.cell.orthogonalize(st.cell.images[im.sym_idx - 1].apply(st.cell.fractionalize(bpos)))
+        dist = atom.pos.dist(bpos)
+        peaks.append((map_val, b.volume, bpos, dist, chain, res, atom))
+
+    if len(peaks) == 0:
+        logger.writeln("No peaks found. Change parameter(s).")
+        return
+        
+    # Print and write coot script
+    peaks.sort(reverse=True, key=lambda x:(abs(x[0]), x[1]))
+    for_coot = []
+    for_df = []
+    for i, p in enumerate(peaks):
+        map_val, volume, mpos, dist, chain, res, atom = p
+        mpos_str = "({: 7.2f},{: 7.2f},{: 7.2f})".format(mpos.x, mpos.y, mpos.z)
+        atom_name = atom.name + ("." + atom.altloc if atom.altloc != "\0" else "")
+        atom_str = "{}/{}/{}".format(chain.name, res.seqid, atom_name)
+        if args.abs_level is None:
+            map_val /= gr_sigma
+        lab_str = "Peak {:4d} value= {: .2e} volume= {:5.1f} pos= {} closest= {:10s} dist= {:.2f}".format(i+1, map_val, volume, mpos_str, atom_str, dist)
+        for_coot.append((lab_str, (mpos.x, mpos.y, mpos.z)))
+        for_df.append((map_val, volume, mpos.x, mpos.y, mpos.z, chain.name, str(res.seqid), res.name, atom_name, dist))
+    df = pandas.DataFrame(for_df, columns=["map_value" if args.abs_level is not None else "sigma_level",
+                                           "volume", "x", "y", "z", "chain", "seqid", "residue", "atom", "dist"])
+    logger.writeln(df.to_string())
+    with open(args.output_prefix + ".json", "w") as ofs:
+        df.to_json(ofs, orient="records", indent=2)
+        logger.writeln("saved: {}".format(ofs.name))
+    coot_out = args.output_prefix + "_coot.py"
+    with open(coot_out, "w") as ofs:
+        ofs.write("""\
+from __future__ import absolute_import, division, print_function
+import gtk
+class coot_serval_map_peak_list:
+  def __init__(self):
+    window = gtk.Window(gtk.WINDOW_TOPLEVEL)
+    window.set_title("Map peaks (Servalcat)")
+    window.set_default_size(600, 600)
+    scrolled_win = gtk.ScrolledWindow()
+    scrolled_win.set_policy(gtk.POLICY_AUTOMATIC, gtk.POLICY_ALWAYS)
+    vbox = gtk.VBox(False, 2)
+    frame_vbox = gtk.VBox(False, 0)
+    frame_vbox.set_border_width(3)
+    self.btns = []
+    self.data = {}
+    self.add_data(frame_vbox)
+    scrolled_win.add_with_viewport(frame_vbox)
+    vbox.pack_start(scrolled_win, True, True, 0)
+    window.add(vbox)
+    window.show_all()
+    self.toggled(self.btns[0], 0)
+
+  def toggled(self, btn, i):
+    if btn.get_active():
+      set_rotation_centre(*self.data[i][1])
+      add_status_bar_text(self.data[i][0])
+
+  def add_data(self, vbox):
+    for i, d in enumerate(self.data):
+      self.btns.append(gtk.RadioButton(None if i == 0 else self.btns[0], d[0]))
+      vbox.pack_start(self.btns[-1], False, False, 0)
+      self.btns[-1].connect('toggled', self.toggled, i)
+
+gui = coot_serval_map_peak_list()
+""".format(for_coot))
+    logger.writeln("\nRun:")
+    logger.writeln("coot --script {}".format(coot_out))
+# map_peaks()
+
+def h_density_analysis(args):
+    #if args.source != "electron":
+    #    raise SystemExit("Only electron source is supported.")
+    model_format = fileio.check_model_format(args.model)
+    st = fileio.read_structure(args.model)
+    if not st[0].has_hydrogen():
+        raise SystemExit("No hydrogen in model.")
+
+    if args.output_prefix is None:
+        args.output_prefix = fileio.splitext(os.path.basename(args.model))[0] + "_hana"
+
+    gr = read_map_and_oversample(map_in=args.map, mtz_in=args.mtz, mtz_labs=args.mtz_labels,
+                                 oversample_pixel=args.oversample_pixel)
+                            
+    if args.abs_level is not None:
+        cutoff = args.abs_level
+    else:
+        cutoff = args.sigma_level * numpy.std(gr) # assuming mean(gr) = 0
+
+    blobs = gemmi.find_blobs_by_flood_fill(gr, cutoff,
+                                           min_volume=args.min_volume, min_score=0)
+    getpos = dict(peak=lambda x: x.peak_pos,
+                  centroid=lambda x: x.centroid)[args.blob_pos]
+
+    peaks = [getpos(b).tolist() for b in blobs]
+    kdtree = scipy.spatial.cKDTree(peaks)
+    found = []
+    n_hydr = 0
+    h_assigned = [0 for _ in range(len(blobs))]
+    st2 = st.clone()
+    for ic, chain in enumerate(st[0]):
+        for ir, res in enumerate(chain):
+            for ia, atom in reversed(list(enumerate(res))):
+                if not atom.is_hydrogen(): continue
+                n_hydr += 1
+                dist, idx = kdtree.query(atom.pos.tolist(), k=1, p=2)
+                map_val = gr.interpolate_value(getpos(blobs[idx]))
+                if dist < args.max_dist and blobs[idx].volume < args.max_volume and map_val > cutoff:
+                    found.append((getpos(blobs[idx]), map_val, dist,  blobs[idx].volume, 
+                                  chain.name, str(res.seqid), res.name,
+                                  atom.name, atom.altloc.replace("\0","")))
+                    h_assigned[idx] = 1
+                else:
+                    del st2[0][ic][ir][ia]
+
+    found.sort(key=lambda x: x[1], reverse=True)
+    logger.writeln("")
+    logger.writeln("Found hydrogen peaks:")
+    logger.writeln("dist map  vol  atom")
+    for _, map_val, dist, volume, chain, resi, resn, atom, alt in found:
+        logger.writeln("{:.2f} {:.2f} {:.2f} {}/{} {}/{}{}".format(dist, map_val, volume,
+                                                                 chain, resn, resi,
+                                                                 atom, "."+alt if alt else ""))
+
+    logger.writeln("")
+    logger.writeln("Result:")
+    logger.writeln(" number of hydrogen in the model  : {}".format(n_hydr))
+    logger.writeln(" number of peaks close to hydrogen: {} ({:.1%})".format(len(found), len(found)/n_hydr))
+    logger.writeln("")
+
+    st_peaks = model.st_from_positions([getpos(b) for b in blobs],
+                                       bs=[gr.interpolate_value(getpos(b)) for b in blobs],
+                                       qs=h_assigned)
+    fileio.write_model(st_peaks, file_name="{}_peaks.mmcif".format(args.output_prefix))
+    logger.writeln(" this file includes peak positions")
+    logger.writeln(" occ=1: hydrogen assigned, occ=0: unassigned.")
+    logger.writeln(" B: density value at {}".format(args.blob_pos))
+    logger.writeln("")
+    
+    fileio.write_model(st2, file_name="{}_h_with_peak{}".format(args.output_prefix, model_format))
+    logger.writeln(" this file is a copy of input model, where hydrogen atoms without peaks are removed.")
+# h_density_analysis()
+
+def fix_link(args):
+    st = fileio.read_structure(args.model)
+    model_format = fileio.check_model_format(args.model)
+    
+    if not args.output:
+        tmp = fileio.splitext(os.path.basename(args.model))[0]
+        args.output = tmp + "_fixlink" + model_format
+    logger.writeln("Output file: {}".format(args.output))
+        
+    args.ligand = sum(args.ligand, []) if args.ligand else []
+    monlib = restraints.load_monomer_library(st,
+                                             monomer_dir=args.monlib,
+                                             cif_files=args.ligand)
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)
+    restraints.find_and_fix_links(st, monlib, bond_margin=args.bond_margin,
+                                  metal_margin=args.metal_margin)
+    fileio.write_model(st, file_name=args.output)
+# fix_link()
+    
+def merge_models(args):
+    logger.writeln("Reading file   1: {}".format(args.models[0]))
+    st = fileio.read_structure(args.models[0])
+    logger.writeln("                  chains {}".format(" ".join([c.name for c in st[0]])))
+
+    for i, f in enumerate(args.models[1:]):
+        logger.writeln("Reading file {:3d}: {}".format(i+2, f))
+        st2 = fileio.read_structure(f)
+        for c in st2[0]:
+            org_id = c.name
+            c2 = st[0].add_chain(c, unique_name=True)
+            if c.name != c2.name:
+                logger.writeln("                  chain {} merged (ID changed to {})".format(c.name, c2.name))
+            else:
+                logger.writeln("                  chain {} merged".format(c.name))
+
+    fileio.write_model(st, file_name=args.output)
+# merge_models()
+
+def merge_dicts(args):
+    fileio.merge_ligand_cif(args.cifs, args.output)
+# merge_dicts()
+
+def geometry(args):
+    if args.ligand: args.ligand = sum(args.ligand, [])
+    if not args.output_prefix: args.output_prefix = fileio.splitext(os.path.basename(args.model))[0] + "_geom"
+    keywords = []
+    if args.keywords or args.keyword_file:
+        if args.keywords: keywords = sum(args.keywords, [])
+        if args.keyword_file: keywords.extend(l for f in sum(args.keyword_file, []) for l in open(f))
+    params = refmac_keywords.parse_keywords(keywords)
+    st = fileio.read_structure(args.model)
+    if args.ignore_h:
+        st.remove_hydrogens()
+    try:
+        monlib = restraints.load_monomer_library(st, monomer_dir=args.monlib, cif_files=args.ligand, 
+                                                 stop_for_unknowns=True, params=params)
+    except RuntimeError as e:
+        raise SystemExit("Error: {}".format(e))
+
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)
+    restraints.find_and_fix_links(st, monlib)
+    try:
+        topo, _ = restraints.prepare_topology(st, monlib, h_change=gemmi.HydrogenChange.NoChange,
+                                              check_hydrogen=True, params=params)
+    except RuntimeError as e:
+        raise SystemExit("Error: {}".format(e))
+    
+    refine_params = RefineParams(st, refine_xyz=True)
+    if args.selection:
+        sel = gemmi.Selection(args.selection)
+        geom_w = [0. for _ in range(st[0].count_atom_sites())]
+        n = 0
+        for chain in sel.chains(st[0]):
+            for res in sel.residues(chain):
+                for atom in sel.atoms(res):
+                    geom_w[atom.serial-1] = 1.
+                    n += 1
+        logger.writeln("Using selection '{}': {} atoms out of {}".format(args.selection, n, len(geom_w)))
+        refine_params.geom_weights[:] = geom_w
+
+    geom = Geom(st, topo, monlib, refine_params,
+                params=params, use_nucleus=args.nucleus)
+    for k in geom.outlier_sigmas: geom.outlier_sigmas[k] = args.sigma
+    geom.setup_nonbonded()
+    ret = geom.show_model_stats()
+    
+    with open(args.output_prefix + "_summary.json", "w") as ofs:
+        ret["summary"].to_json(ofs, indent=2)
+        logger.writeln("saved: {}".format(ofs.name))
+    with open(args.output_prefix + "_outliers.json", "w") as ofs:
+        for k in ret["outliers"]:
+            ret["outliers"][k] = ret["outliers"][k].to_dict(orient="records")
+        json.dump(ret["outliers"], ofs, indent=2)
+        logger.writeln("saved: {}".format(ofs.name))
+
+    if args.dump_all: # for debug, unfinished
+        dd = {"bonds": [], "vdw": []}
+        for t in geom.geom.bonds:
+            dd["bonds"].append({"atoms":[str(geom.lookup[x]) for x in t.atoms],
+                                "ideals": [{"ideal":x.value, "sigma":x.sigma} for x in t.values],
+                                "alpha": t.alpha, "type": t.type,
+                                "sym_idx": t.sym_idx, "pbc_shift": t.pbc_shift})
+        for t in geom.geom.vdws:
+            dd["vdw"].append({"atoms":[str(geom.lookup[x]) for x in t.atoms],
+                              "ideal": t.value, "sigma": t.sigma,
+                              "type": t.type, "sym_idx": t.sym_idx, "pbc_shift": t.pbc_shift})
+        with open(args.output_prefix + "_restraints.json", "w") as ofs:
+            json.dump(dd, ofs)
+            logger.writeln("saved: {}".format(ofs.name))
+    if args.check_skew:
+        logger.writeln("\nChecking skewness of bond length deviation")
+        # better to ignore hydrogen
+        tab = geom.geom.reporting.get_bond_outliers(use_nucleus=geom.use_nucleus, min_z=0)
+        for i in range(2):
+            tab[f"atom{i+1}"] = [str(geom.lookup[r.atoms[i]]) for r in tab["restr"]]
+        del tab["restr"]
+        df = pandas.DataFrame(tab)
+        df["dev"] = df["value"] - df["ideal"]
+        df = df.reindex(df.dev.abs().sort_values(ascending=False).index)
+        logger.writeln("Bond length deviations:")
+        logger.writeln(df.to_string(max_rows=20))
+        q1, q2, q3 = numpy.percentile(df["dev"], [25, 50, 75])
+        sk2 = (q1 + q3 - 2 * q2) / (q3 - q1)
+        logger.writeln("bond_dev_median= {:.6f}".format(q2))
+        logger.writeln("bond_dev_skew=   {:.4f}".format(df["dev"].skew()))
+        logger.writeln("bond_dev_sk2=    {:.4f}".format(sk2))
+        with open(args.output_prefix + "_bond_dev.html", "w") as ofs:
+            ofs.write("""\
+<html>
+<head>
+ <meta charset="utf-8" />
+ <script src="https://cdn.plot.ly/plotly-2.20.0.min.js" charset="utf-8"></script>
+</head>
+<body>
+ <div id="hist"></div>
+ <script>
+  var trace = {
+   x: %s,
+   type: 'histogram'
+  };
+  var layout = {
+   title: "median: %.4f, sk2: %.4f",
+   xaxis: {title: "bond distance - ideal"},
+   yaxis: {title: "count"},
+   shapes: [{
+    type: 'line',
+    yref: 'paper',
+    x0: 0, y0: 0,
+    x1: 0, y1: 1}]
+  };
+  target = document.getElementById('hist');
+  Plotly.newPlot(target, [trace], layout);
+ </script>
+</body>
+</html>
+""" % (str(list(df.dev)), q2, sk2))
+            logger.writeln("check histogram: {}".format(ofs.name))
+
+    # Note that this modifies st
+    if args.per_atom_score_as_b:
+        model_format = fileio.check_model_format(args.model)
+        peratom = geom.geom.reporting.per_atom_score(len(geom.atoms), geom.use_nucleus, "mean")
+        for i, score in enumerate(peratom["total"]):
+            geom.atoms[i].b_iso = score
+        fileio.write_model(st, file_name="{}_per_atom_score{}".format(args.output_prefix, model_format))
+# geometry()
+
+def compare_conf(args):
+    def angle_abs_diff(a, b, full=360.):
+        # from gemmi/math.hpp
+        d = abs(a - b)
+        if d > full:
+            d -= numpy.floor(d / full) * full
+        return min(d, full - d)
+    # angle_abs_diff()
+    
+    if args.ligand: args.ligand = sum(args.ligand, [])
+    st = None
+    for i, f in enumerate(args.models):
+        tmp = fileio.read_structure(f)
+        if len(args.models) > 1:
+            for chain in tmp[0]:
+                chain.name = f"{i+1}_{chain.name}"
+        if i == 0:
+            st = tmp
+        else:
+            for chain in tmp[0]:
+                st[0].add_chain(chain)
+    try:
+        monlib = restraints.load_monomer_library(st, monomer_dir=args.monlib, cif_files=args.ligand, 
+                                                 stop_for_unknowns=True)
+    except RuntimeError as e:
+        raise SystemExit(f"Error: {e}")
+
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)
+    try:
+        topo, _ = restraints.prepare_topology(st, monlib, h_change=gemmi.HydrogenChange.NoChange,
+                                              check_hydrogen=False)
+    except RuntimeError as e:
+        raise SystemExit(f"Error: {e}")
+    ncslist = restraints.prepare_ncs_restraints(st)
+    lookup = {x.atom: x for x in st[0].all()}
+    ptypes = {x.name: x.polymer_type for x in st.entities}
+    resn_lookup = {(chain.name, res.seqid): res.name for chain in st[0] for res in chain}
+    confs = {}
+    for t in topo.torsions:
+        cra = lookup[t.atoms[0]]
+        ptype = ptypes[cra.residue.entity_id]
+        is_peptide = ptype in (gemmi.PolymerType.PeptideL, gemmi.PolymerType.PeptideD)
+        is_peptide_tors = t.restr.label.startswith("chi") or t.restr.label in ("omega", "phi", "psi")
+        is_na = ptype in (gemmi.PolymerType.Dna, gemmi.PolymerType.Rna, gemmi.PolymerType.DnaRnaHybrid)
+        is_na_tors = t.restr.label in ("C2e-chi", "alpha", "beta", "gamma", "C2e-nyu0", "epsilon", "zeta")
+        if (is_peptide and is_peptide_tors) or (is_na and is_na_tors):
+            confs.setdefault(cra.chain.name, {}).setdefault(cra.residue.seqid, {})[t.restr.label] = numpy.rad2deg(t.calculate())
+    fulls = {("ARG", "chi5"): 180., ("TYR", "chi2"): 180., ("PHE", "chi2"): 180., ("ASP", "chi2"): 180., ("GLU", "chi3"): 180.}
+    ret = []
+    for_coot = []
+    for ncs in ncslist.ncss:
+        c1, c2 = ncs.chains
+        if args.same_chain and len(args.models) > 1 and c1[c1.index("_"):] != c2[c2.index("_"):]:
+            continue
+        for s1, s2 in ncs.seqids:
+            if c1 in confs and s1 in confs[c1] and c2 in confs and s2 in confs[c2]:
+                conf1, conf2 = confs[c1][s1], confs[c2][s2]
+                resn = resn_lookup[(c1, s1)]
+                for t in conf1:
+                    if t in conf2:
+                        d = float(angle_abs_diff(conf1[t], conf2[t], fulls.get((resn, t), 360.)))
+                        ret.append((c1, s1, c2, s2, resn, t, conf1[t], conf2[t], d))
+                        if d > args.min_diff:
+                            for_coot.append((c1, s1.num, c2, s2.num, resn, t, d))
+    df = pandas.DataFrame(ret, columns=["chain_1", "seq_1", "chain_2", "seq_2", "resn", "label", "conf_1", "conf_2", "diff"])
+    df.sort_values("diff", ascending=False, inplace=True)
+    logger.writeln(f"\nList of torsion angle differences (>{args.min_diff})")
+    logger.writeln(df[df["diff"] > args.min_diff].to_string(index=False))
+
+    for_coot.sort(key=lambda x:-x[-1])
+    coot_out = args.output_prefix + "_coot.py"
+    with open(coot_out, "w") as ofs:
+        # https://python-gtk-3-tutorial.readthedocs.io/en/latest/treeview.html
+        ofs.write("""\
+from __future__ import absolute_import, division, print_function
+import re
+import gtk
+class coot_serval_conf_list:
+  def __init__(self):
+    window = gtk.Window(gtk.WINDOW_TOPLEVEL)
+    window.set_title("Different conformations (Servalcat)")
+    window.set_default_size(600, 600)
+    scrolled_win = gtk.ScrolledWindow()
+    scrolled_win.set_policy(gtk.POLICY_AUTOMATIC, gtk.POLICY_ALWAYS)
+    vbox = gtk.VBox(False, 2)
+    self.liststore = gtk.ListStore(str, int, str, int, str, str, float)
+    self.filter = self.liststore.filter_new()
+    self.treeview = gtk.TreeView(model=self.filter)
+    for i, column_title in enumerate(["chain_1", "seq_1", "chain_2", "seq_2", "resn", "label", "diff"]):
+      renderer = gtk.CellRendererText()
+      column = gtk.TreeViewColumn(column_title, renderer, text=i)
+      self.treeview.append_column(column)
+    self.data = {}
+    self.add_data()
+    scrolled_win.add_with_viewport(self.treeview)  # add?
+    vbox.pack_start(scrolled_win, True, True, 0)
+    window.add(vbox)
+    window.show_all()
+    self.treeview.connect("row-activated", self.on_row_activated)
+
+  def on_row_activated(self, treeview, path, column):
+    assert len(path) == 1
+    col_idx = [i for i, c in enumerate(treeview.get_columns()) if column == c][0]
+    row = self.liststore[path[0]]
+    if col_idx < 2:
+      chain, resi = row[0], row[1]
+    elif col_idx < 4:
+      chain, resi = row[2], row[3]
+    else:
+      return
+    if re.search("^[0-9]+_[0-9A-Za-z]", chain):
+      chain = chain[chain.index("_")+1:]
+    imol = active_atom_spec()[1][0]
+    for name in (" CA ", " C1'"):
+      a = get_atom(imol, chain, resi, "", name)
+      if a:
+        set_rotation_center(*a[2])
+        break
+
+  def add_data(self):
+    for i, d in enumerate(self.data):
+      self.liststore.append(d)
+
+gui = coot_serval_conf_list()
+""".format(for_coot))
+    logger.writeln("\nRun:")
+    logger.writeln(f"coot --script {coot_out}")
+# compare_conf()
+
+def adp_stats(args):
+    if not args.output_prefix: args.output_prefix = fileio.splitext(os.path.basename(args.model))[0] + "_adp"
+    st = fileio.read_structure(args.model)
+    model.adp_analysis(st)
+    b_all = [cra.atom.b_iso for cra in st[0].all() if cra.atom.occ > 0]
+    
+    # bin width from Freedman–Diaconis rule
+    qs = numpy.quantile(b_all, [0, 0.25, 0.75, 1])
+    bin_h = 2 * (qs[2] - qs[1]) / len(b_all)**(1/3.)
+    
+    # for plotly
+    traces = []
+    traces.append("x: [%s], type: 'histogram', name: 'All', xbins: {size: %f}"
+                  % (",".join("%.2f"%x for x in b_all), bin_h))
+    if len(st[0]) > 1:
+        b_chain = {}
+        for c in st[0]:
+            b_chain.setdefault(c.name, []).extend(a.b_iso for r in c for a in r if a.occ > 0)
+        for c in b_chain:
+            bs = ",".join("%.2f" % x for x in b_chain[c])
+            traces.append("x: [%s], type: 'histogram', name: 'Chain %s'" % (bs, c))
+    with open(args.output_prefix + "_hist.html", "w") as ofs:
+        ofs.write("""\
+<html>
+<head>
+ <meta charset="utf-8" />
+ <script src="https://cdn.plot.ly/plotly-2.20.0.min.js" charset="utf-8"></script>
+</head>
+<body>
+ <div id="hist"></div>
+ <script>
+""")
+        for i, t in enumerate(traces):
+            ofs.write("var trace%d = {%s};\n" % (i+1, t))
+        ofs.write("""\
+  var layout = {
+   title: "isotropic B histogram",
+   xaxis: {title: "B"},
+   yaxis: {title: "count"},
+   barmode: "stack"
+  };
+  target = document.getElementById('hist');
+  Plotly.newPlot(target, [%s], layout);
+ </script>
+</body>
+</html>
+""" % (",".join("trace%d" % (i+1) for i in range(len(traces)))))
+        logger.writeln("check histogram: {}".format(ofs.name))
+# adp_stats()
+
+def show_power(args):
+    maps_in = []
+    if args.map:
+        print(args.map)
+        print(sum(args.map, []))
+        maps_in = [(f,) for f in sum(args.map, [])]
+        
+    if args.halfmaps:
+        args.halfmaps = sum(args.halfmaps, [])
+        if len(args.halfmaps)%2 != 0:
+            raise RuntimeError("Number of half maps is not even.")
+        maps_in.extend([(args.halfmaps[2*i],args.halfmaps[2*i+1]) for i in range(len(args.halfmaps)//2)])
+        
+    if args.mask:
+        mask = fileio.read_ccp4_map(args.mask)[0]
+    else:
+        mask = None
+
+    hkldata = None
+    labs = []
+    for mapin in maps_in: # TODO rewrite in faster way
+        ms = [fileio.read_ccp4_map(f) for f in mapin]
+        d_min = args.resolution
+        if d_min is None:
+            d_min = maps.nyquist_resolution(ms[0][0])
+            logger.writeln("WARNING: --resolution is not specified. Using Nyquist resolution: {:.2f}".format(d_min))
+        tmp = maps.mask_and_fft_maps(ms, d_min, mask)
+        labs.append("F{:02d}".format(len(labs)+1))
+        tmp.df.rename(columns=dict(FP=labs[-1]), inplace=True)
+        if hkldata is None:
+            hkldata = tmp
+        else:
+            if hkldata.cell.parameters != tmp.cell.parameters: raise RuntimeError("Different unit cell!")
+            hkldata.merge(tmp.df[["H","K","L",labs[-1]]])
+
+    if not labs:
+        raise SystemExit("No map files given. Exiting.")
+            
+    hkldata.setup_relion_binning("stat")
+
+    ofs = open(args.output_prefix+".log", "w")
+    ofs.write("Input:\n")
+    for i in range(len(maps_in)):
+        ofs.write("{} from {}\n".format(labs[i], " ".join(maps_in[i])))
+    ofs.write("\n")
+    
+    ofs.write("""$TABLE: Power spectrum :
+$GRAPHS
+: log10(Mn(|F|^2)) :A:1,{}:
+$$
+1/resol^2 n d_max d_min {}
+$$
+$$
+""".format(",".join([str(i+5) for i in range(len(labs))]), " ".join(labs)))
+    print(hkldata.df)
+    abssqr = dict((lab, numpy.abs(hkldata.df[lab].to_numpy())**2) for lab in labs)
+    for i_bin, idxes in hkldata.binned("stat"):
+        bin_d_min = hkldata.binned_df["stat"].d_min[i_bin]
+        bin_d_max = hkldata.binned_df["stat"].d_max[i_bin]
+        ofs.write("{:.4f} {:7d} {:7.3f} {:7.3f}".format(1/bin_d_min**2, len(idxes), bin_d_max, bin_d_min,))
+        for lab in labs:
+            pwr = numpy.log10(numpy.average(abssqr[lab][idxes]))
+            ofs.write(" {:.4e}".format(pwr))
+        ofs.write("\n")
+    ofs.write("$$\n")
+    ofs.close()
+# show_power()
+
+def fcalc(args):
+    if (args.auto_box_with_padding, args.cell).count(None) == 0:
+        raise SystemExit("Error: you cannot specify both --auto_box_with_padding and --cell")
+    
+    if args.ligand: args.ligand = sum(args.ligand, [])
+    if not args.output_prefix: args.output_prefix = "{}_fcalc_{}".format(fileio.splitext(os.path.basename(args.model))[0], args.source)
+
+    st = fileio.read_structure(args.model)
+    ccu = model.CustomCoefUtil()
+    if not args.keep_charges:
+        model.remove_charge([st])
+    if args.source == "custom":
+        ccu.read_from_cif(st, args.model)
+        ccu.show_info()
+        ccu.set_coeffs(st)
+    else:
+        model.check_atomsf([st], args.source)
+    if not args.no_expand_ncs:
+        model.expand_ncs(st)    
+
+    if args.cell is not None:
+        st.cell = gemmi.UnitCell(*args.cell)
+    elif args.auto_box_with_padding is not None:
+        st.cell = model.box_from_model(st[0], args.auto_box_with_padding)
+        st.spacegroup_hm = "P 1"
+        logger.writeln("Box size from the model with padding of {}: {}".format(args.auto_box_with_padding, st.cell.parameters))
+        
+    if not st.cell.is_crystal():
+        raise SystemExit("ERROR: No unit cell information. Give --cell or --auto_box_with_padding.")
+
+    if args.source=="electron" and st[0].has_hydrogen():
+        monlib = restraints.load_monomer_library(st, monomer_dir=args.monlib, cif_files=args.ligand, 
+                                                 stop_for_unknowns=False)
+    else:
+        monlib = None
+
+    if args.method == "fft":
+        fc_asu = model.calc_fc_fft(st, args.resolution, cutoff=args.cutoff, rate=args.rate,
+                                   mott_bethe=args.source=="electron",
+                                   monlib=monlib, source=args.source)
+    else:
+        fc_asu = model.calc_fc_direct(st, args.resolution, source=args.source,
+                                      mott_bethe=args.source=="electron", monlib=monlib)
+
+    hkldata = hkl.hkldata_from_asu_data(fc_asu, "FC")
+    if args.as_intensity:
+        hkldata.df["IC"] = numpy.abs(hkldata.df.FC)**2
+        labout = ["IC"]
+        if args.add_dummy_sigma:
+            hkldata.df["SIGIC"] = 1.
+            labout.append("SIGIC")
+    else:
+        labout = ["FC"]
+        if args.add_dummy_sigma:
+            hkldata.df["SIGFC"] = 1.
+            labout.append("SIGFC")
+
+    hkldata.write_mtz(args.output_prefix+".mtz", labout, types=dict(IC="J", SIGIC="Q", SIGFC="Q"))
+# fcalc()
+
+def nemap(args):
+    from servalcat.spa import fofc
+
+    if (args.trim or args.trim_mtz) and args.mask is None:
+        raise SystemExit("\nError: You need to give --mask as you requested --trim or --trim_mtz.\n")
+    
+    if args.mask:
+        mask = fileio.read_ccp4_map(args.mask)[0]
+    else:
+        mask = None
+
+    halfmaps = fileio.read_halfmaps(args.halfmaps, pixel_size=args.pixel_size)
+    if args.resolution is None:
+        args.resolution = maps.nyquist_resolution(halfmaps[0][0])
+        logger.writeln("WARNING: --resolution is not specified. Using Nyquist resolution: {:.2f}".format(args.resolution))
+
+    d_min = args.resolution
+    if args.local_fourier_weighting_with > 0:
+        d_min = 1 / (args.local_fourier_weighting_with + 1 / d_min)
+        logger.writeln("adjusting d_min= {:.2f} for local correlation".format(d_min))
+    hkldata = maps.mask_and_fft_maps(halfmaps, d_min, mask)
+
+    if args.local_fourier_weighting_with > 0:
+        asu1 = hkldata.as_asu_data("F_map1")
+        asu2 = hkldata.as_asu_data("F_map2")
+        size = asu1.get_size_for_hkl(sample_rate=3)
+        logger.writeln("using grid {}".format(size))
+        gr1 = asu1.get_f_phi_on_grid(size)
+        gr2 = asu2.get_f_phi_on_grid(size)
+        kernel = ext.hard_sphere_kernel_recgrid(size, asu1.unit_cell, args.local_fourier_weighting_with)
+        cc = maps.local_cc(gr1, gr2, kernel.array.real, method="simple")
+        cc.array[cc.array < 0] = 0 # negative cc cannot be used anyway
+        cc.array[:] = 2 * cc.array.real / (1 + cc.array.real) # to full map cc
+        hkldata.df["cc"] = numpy.real(cc.get_value_by_hkl(hkldata.miller_array()))
+        grf = type(gr1)((gr1.array + gr2.array) / 2, gr1.unit_cell, gr1.spacegroup)
+        var_f = maps.local_var(grf, kernel.array.real, method="simple")
+        hkldata.df["var_f"] = numpy.real(var_f.get_value_by_hkl(hkldata.miller_array()))
+        if args.B is not None:
+            k2_l = numpy.exp(-args.B / hkldata.d_spacings()**2 / 2)
+            hkldata.df.cc = k2_l * hkldata.df.cc / (1 + (k2_l - 1) * hkldata.df.cc)
+        hkldata.df["FWT"] = hkldata.df.FP * numpy.sqrt(hkldata.df.cc / hkldata.df.var_f)
+        hkldata.df["kernel"] = numpy.real(kernel.get_value_by_hkl(hkldata.miller_array()))
+        hkldata.write_mtz(args.output_prefix+"_cc.mtz", ["cc", "kernel"])
+        hkldata = hkldata.copy(d_min=args.resolution)
+        map_labs = ["FWT"]
+    else:
+        hkldata.setup_relion_binning("ml")
+        maps.calc_noise_var_from_halfmaps(hkldata)
+        map_labs = fofc.calc_maps(hkldata, B=args.B, has_halfmaps=True, half1_only=args.half1_only,
+                                  no_fsc_weights=args.no_fsc_weights, sharpening_b=args.sharpening_b)
+    fofc.write_files(hkldata, map_labs, grid_start=halfmaps[0][1], stats_str=None,
+                     mask=mask, output_prefix=args.output_prefix,
+                     trim_map=args.trim, trim_mtz=args.trim_mtz)
+# nemap()
+
+def blur(args):
+    if args.output_prefix is None:
+        args.output_prefix = fileio.splitext(os.path.basename(args.hklin))[0]
+    
+    if fileio.is_mmhkl_file(args.hklin):
+        mtz = fileio.read_mmhkl(args.hklin)
+        hkl.blur_mtz(mtz, args.B)
+        suffix = ("_blur" if args.B > 0 else "_sharpen") + "_{:.2f}.mtz".format(abs(args.B))
+        mtz.write_to_file(args.output_prefix+suffix)
+        logger.writeln("Written: {}".format(args.output_prefix+suffix))
+    else:
+        raise SystemExit("ERROR: Unsupported file type: {}".format(args.hklin))
+# blur()
+
+def mask_from_model(args):
+    st = fileio.read_structure(args.model) # TODO option to (or not to) expand NCS
+    if args.selection:
+        gemmi.Selection(args.selection).remove_not_selected(st)
+    gr, grid_start, _ = fileio.read_ccp4_map(args.map, header_only=True)
+    mask = maps.mask_from_model(st, args.radius, soft_edge=args.soft_edge, grid=gr)
+    maps.write_ccp4_map(args.output, mask, grid_start=grid_start)
+# mask_from_model()
+
+def applymask(args):
+    if args.output_prefix is None:
+        args.output_prefix = fileio.splitext(os.path.basename(args.map))[0] + "_masked"
+
+    grid, grid_start, _ = fileio.read_ccp4_map(args.map)
+    mask = fileio.read_ccp4_map(args.mask)[0]
+    logger.writeln("Applying mask")
+    logger.writeln(" mask min: {:.3f} max: {:.3f}".format(numpy.min(mask), numpy.max(mask)))
+    grid.array[:] *= mask.array
+
+    if args.normalize:
+        masked = grid.array[mask.array>args.mask_cutoff]
+        masked_mean = numpy.average(masked)
+        masked_std = numpy.std(masked)
+        logger.writeln("Normalizing map values within mask")
+        logger.writeln(" masked volume: {} mean: {:.3e} sd: {:.3e}".format(len(masked), masked_mean, masked_std))
+        grid.array[:] = (grid.array - masked_mean) / masked_std
+
+    maps.write_ccp4_map(args.output_prefix+".mrc", grid,
+                        grid_start=grid_start,
+                        mask_for_extent=mask.array if args.trim else None,
+                        mask_threshold=args.mask_cutoff)
+# applymask()
+
+def map2mtz(args):
+    if args.output is None:
+        args.output = fileio.splitext(os.path.basename(args.map))[0] + "_fft.mtz"
+    grid, grid_start, grid_shape = fileio.read_ccp4_map(args.map)
+    if args.resolution is None:
+        args.resolution = maps.nyquist_resolution(grid)
+        logger.writeln("WARNING: --resolution is not specified. Using Nyquist resolution: {:.2f}".format(args.resolution))
+
+    if grid_start != (0,0,0) or grid.shape != tuple(grid_shape):
+        # If only subregion of whole grid in map, unit cell needs to be re-defined.
+        if grid.shape != tuple(grid_shape):
+            new_abc = [grid.unit_cell.parameters[i] * grid_shape[i] / grid.shape[i] for i in range(3)]
+            cell = gemmi.UnitCell(*new_abc, *grid.unit_cell.parameters[3:])
+            logger.writeln("Changing unit cell to {}".format(cell.parameters))
+        else:
+            cell = grid.unit_cell
+        grid = gemmi.FloatGrid(grid.get_subarray(grid_start, grid_shape),
+                               cell, grid.spacegroup)
+    
+    f_grid = gemmi.transform_map_to_f_phi(grid)
+    asudata = f_grid.prepare_asu_data(dmin=args.resolution, with_000=True)
+    hkldata = hkl.hkldata_from_asu_data(asudata, "F")
+    if grid_start != (0,0,0):
+        shifts = grid.get_position(*grid_start)
+        hkldata.translate("F", shifts)
+        logger.writeln("Applying phase shift with translation {}".format(shifts.tolist()))
+    hkldata.write_mtz(args.output, ["F"])
+# map2mtz()
+
+def sm2mm(args):
+    if args.output_prefix is None:
+        args.output_prefix = os.path.basename(fileio.splitext(args.files[0])[0])
+    st, mtz = fileio.read_small_molecule_files(args.files)
+    if st is not None:
+        fileio.write_model(st, prefix=args.output_prefix, pdb=True, cif=True)
+    if mtz is not None:
+        mtz_out = args.output_prefix + ".mtz"
+        logger.writeln("Writing MTZ file: {}".format(mtz_out))
+        mtz.write_to_file(mtz_out)
+# sm2mm()
+
+def mm2ins(args):
+    if args.output is None:
+        args.output = os.path.basename(fileio.splitext(args.model)[0]) + ".ins"
+    st = fileio.read_structure(args.model)
+    sg = st.find_spacegroup()
+    elems = [cra.atom.element.name for cra in st[0].all()]
+    counts = {x:elems.count(x) for x in set(elems)}
+    elems = sorted(counts)
+    cell = st.cell
+    if args.hklin:
+        mtz = fileio.read_mmhkl(args.hklin)
+        cell = mtz.cell
+        wavelength = next((x.wavelength for x in mtz.datasets if x==x and x.wavelength > 0), None)
+    else:
+        mtz = None
+        wavelength = None
+
+    latt = dict(P=1, I=2, R=3, F=4, A=5, B=6, C=7).get(sg.centring_type())
+    if not sg.is_centrosymmetric():
+        latt *= -1
+        
+    with open(args.output, "w") as ofs:
+        ofs.write(f"TITL {os.path.basename(args.model)} in {sg.xhm()}\n")
+        ofs.write(f"CELL {wavelength if wavelength else '????'} ")
+        ofs.write(" ".join(str(x) for x in cell.parameters) + "\n")
+        ofs.write("ZERR 1 0 0 0 0 0 0\n")
+        ofs.write(f"LATT {latt}\n")
+        for op in sg.operations().sym_ops[1:]: # the first is identity
+            ofs.write(f"SYMM {op.triplet('X')}\n")
+        ofs.write(f"SFAC {' '.join(elems)}\n")
+        ofs.write(f"UNIT {' '.join(str(int(counts[x])) for x in elems)}\n\n")
+        ofs.write("""\
+L.S. 10
+ACTA
+LIST 6
+MORE -1\n\n""")
+        for cra in st[0].all():
+            frac = st.cell.fractionalize(cra.atom.pos)
+            u_iso = model.b_to_u * cra.atom.b_iso
+            if cra.atom.is_hydrogen():
+                u_iso = -1.2
+            ofs.write(f"{cra.atom.name} {elems.index(cra.atom.element.name)+1} {frac.x:.6f} {frac.y:.6f} {frac.z:.6f} {10+cra.atom.occ} {u_iso:.5f}\n")
+        logger.writeln(f"Written: {args.output}")
+# mm2ins()
+
+def seq(args):
+    wrap_width = 100
+    seqs = []
+    if args.seq:
+        args.seq = sum(args.seq, [])
+        for sf in args.seq:
+            seqs.extend(fileio.read_sequence_file(sf))
+        
+    sc = gemmi.AlignmentScoring()
+    sc.match, sc.mismatch, sc.gapo, sc.gape, sc.good_gapo, sc.bad_gapo = args.scoring
+    
+    st = fileio.read_structure(args.model) # TODO option to (or not to) expand NCS
+    model.setup_entities(st, clear=True, force_subchain_names=True, overwrite_entity_type=True)    
+    for chain in st[0]:
+        p = chain.get_polymer()
+        if not p: continue
+        p_type = p.check_polymer_type()
+        if p_type in (gemmi.PolymerType.SaccharideD, gemmi.PolymerType.SaccharideL): continue
+        p_seq = gemmi.one_letter_code(p.extract_sequence())
+        results = []
+        for name, seq in seqs:
+            # what if DnaRnaHybrid?
+            kind = {gemmi.PolymerType.Dna: gemmi.ResidueKind.DNA,
+                    gemmi.PolymerType.Rna: gemmi.ResidueKind.RNA}.get(p_type, gemmi.ResidueKind.AA)
+            s = [gemmi.expand_one_letter(x, kind) for x in seq]
+            if None in s: continue
+            #als = [gemmi.align_sequence_to_polymer(s, p, p_type, gemmi.AlignmentScoring(x)) for x in ("s", "p")]
+            #results.append([name, max(als, key=lambda x: x.match_count), seq])
+            results.append([name, gemmi.align_sequence_to_polymer(s, p, p_type, sc), seq])
+
+        if results:
+            logger.writeln("Chain: {}".format(chain.name))
+            logger.writeln(" polymer type: {}".format(str(p_type).replace("PolymerType.", "")))
+            name, al, s1 = max(results, key=lambda x: (x[1].match_count, x[1].score))
+            logger.writeln(" match: {}".format(name))
+            logger.writeln(" aligned: {}".format(al.match_count))
+            logger.writeln(" score: {}".format(al.score))
+            p1, p2 = al.add_gaps(s1, 1), al.add_gaps(p_seq, 2)
+            unkseq = [x.start() for x in re.finditer(r"\-", p1)]
+            mismatches = [x.start() for x in re.finditer(r"\.", al.match_string)]
+            if mismatches or unkseq:
+                idxes = {x.start(): i for i, x in enumerate(re.finditer("[^-]", p2))}
+                seqnums = [str(x.seqid) for x in p]
+                if mismatches:
+                    logger.write(" mismatches: ")
+                    logger.writeln(", ".join("{}({}>{})".format(seqnums[idxes[i]], p1[i], p2[i]) for i in mismatches))
+                if unkseq:
+                    logger.write(" unknown sequence: ")
+                    logger.writeln(", ".join("{}({})".format(seqnums[idxes[i]], p2[i]) for i in unkseq))
+
+            logger.writeln("")
+            for i in range(0, len(p1), wrap_width):
+                logger.writeln(" seq.  {}".format(p1[i:i+wrap_width]))
+                logger.writeln("       {}".format(al.match_string[i:i+wrap_width]))
+                logger.writeln(" model {}\n".format(p2[i:i+wrap_width]))
+        else:
+            logger.writeln("> Chain: {}".format(chain.name))
+            logger.writeln(gemmi.one_letter_code(p.extract_sequence()))
+            logger.writeln("")
+# seq()
+
+def dnarna(args):
+    import scipy.spatial.transform
+    rna_res = {"A":"DA", "G":"DG", "C":"DC", "U":"DT"}
+    dna_res = {"DA":"A", "DG":"G", "DC":"C", "DT":"U"}
+    if args.chains: args.chains = sum(args.chains, [])
+    model_format = fileio.check_model_format(args.model)
+    if not args.output:
+        args.output = fileio.splitext(os.path.basename(args.model))[0] + "_conv" + model_format
+    st = fileio.read_structure(args.model)
+    if st[0].has_hydrogen():
+        logger.writeln("Hydrogen atoms are detected. I cannot take care of them, so I will remove them.")
+        st.remove_hydrogens()
+    for chain in st[0]:
+        if args.chains and chain.name not in args.chains:
+            continue
+        for res in chain:
+            alt = "*" # XXX
+            if res.name in rna_res and args.to_dna:
+                logger.writeln(f"Changing {chain.name}/{res.seqid} {res.name} to DNA")
+                res.name = rna_res[res.name]
+                res.remove_atom("O2'", alt)
+                if res.name == "DT":
+                    C4 = res.find_atom("C4", alt)
+                    C5 = res.find_atom("C5", alt)
+                    C6 = res.find_atom("C6", alt)
+                    v1 = C5.pos - C4.pos
+                    v2 = C5.pos - C6.pos
+                    v = v1 + v2
+                    res.add_atom(C5)
+                    res[-1].name = "C7"
+                    res[-1].pos = C5.pos + v / v.length() * 1.5
+            elif res.name in dna_res and args.to_rna:
+                logger.writeln(f"Changing {chain.name}/{res.seqid} {res.name} to RNA")
+                res.name = dna_res[res.name]
+                C1p = numpy.array(res.find_atom("C1'", alt).pos.tolist())
+                C2p = numpy.array(res.find_atom("C2'", alt).pos.tolist())
+                C3p = numpy.array(res.find_atom("C3'", alt).pos.tolist())
+                rotvec = C2p - C3p
+                rotvec /= numpy.linalg.norm(rotvec)
+                r = scipy.spatial.transform.Rotation.from_rotvec(-rotvec * 120,
+                                                                 degrees=True)
+                rotated = r.apply(C1p - C2p)
+                rotated *= 1.411 / numpy.linalg.norm(rotated)
+                res.add_atom(res.find_atom("O3'", alt))
+                res[-1].name = "O2'"
+                res[-1].pos.fromlist(C2p + rotated)
+                if res.name == "U":
+                    res.remove_atom("C7", alt)
+    fileio.write_model(st, file_name=args.output)
+# dnarna()
+
+def show(args):
+    for filename in args.files:
+        ext = fileio.splitext(filename)[1]
+        if ext in (".mrc", ".ccp4", ".map"):
+            fileio.read_ccp4_map(filename)
+            logger.writeln("\n")
+# show()
+
+def json2csv(args):
+    if not args.output_prefix:
+        args.output_prefix = fileio.splitext(os.path.basename(args.json))[0]
+        
+    df = pandas.read_json(args.json)
+    df.to_csv(args.output_prefix+".csv", index=False)
+    logger.writeln("Output: {}".format(args.output_prefix+".csv"))
+# json2csv()
+
+def main(args):
+    comms = dict(show=show,
+                 json2csv=json2csv,
+                 symmodel=symmodel,
+                 helical_biomt=helical_biomt,
+                 expand=symexpand,
+                 h_add=h_add,
+                 add_op3=add_op3,
+                 map_peaks=map_peaks,
+                 h_density=h_density_analysis,
+                 fix_link=fix_link,
+                 merge_models=merge_models,
+                 merge_dicts=merge_dicts,
+                 geom=geometry,
+                 conf=compare_conf,
+                 adp=adp_stats,
+                 power=show_power,
+                 fcalc=fcalc,
+                 nemap=nemap,
+                 blur=blur,
+                 mask_from_model=mask_from_model,
+                 applymask=applymask,
+                 map2mtz=map2mtz,
+                 sm2mm=sm2mm,
+                 mm2ins=mm2ins,
+                 seq=seq,
+                 dnarna=dnarna)
+    
+    com = args.subcommand
+    f = comms.get(com)
+    if f:
+        return f(args)
+    else:
+        raise SystemExit("Unknown subcommand: {}".format(com))
+# main()
+
+if __name__ == "__main__":
+    import sys
+    args = parse_args(sys.argv[1:])
+    main(args)