PGAP2 2.0__tar.gz → 2.3__tar.gz

{pgap2-2.0 → pgap2-2.3/PGAP2.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PGAP2
-Version: 2.0
+Version: 2.3
 Summary: PGAP2: a comprehensive pan-genome analysis pipeline for prokaryotic genomes
 Author-email: bucongfan <bucf@big.ac.cn>
 Maintainer-email: bucongfan <bucf@big.ac.cn>
@@ -28,6 +28,8 @@ Requires-Dist: ete3==3.1.3
 Requires-Dist: svgwrite==1.4.3
 Requires-Dist: miniFasta==3.0.2
 Requires-Dist: tajimas_d==2.0.2
+Requires-Dist: sqlalchemy>=2.0
+Requires-Dist: igraph
 Dynamic: license-file
 
 ![image](https://github.com/bucongfan/PGAP2/blob/main/pgap2.logo.png)

{pgap2-2.0 → pgap2-2.3}/PGAP2.egg-info/SOURCES.txt

@@ -11,12 +11,14 @@ PGAP2.egg-info/top_level.txt
 pgap2/__init__.py
 pgap2/main.py
 pgap2/cli/__init__.py
+pgap2/cli/neighborhood.py
 pgap2/cli/partition.py
 pgap2/cli/partition_add.py
 pgap2/cli/partition_merge.py
 pgap2/cli/partition_minus.py
 pgap2/cli/postprocess.py
 pgap2/cli/preprocess.py
+pgap2/cli/sqlite_cmd.py
 pgap2/dependencies/draw_post_profile.r
 pgap2/dependencies/draw_post_stat.r
 pgap2/dependencies/draw_prep.r
@@ -31,10 +33,26 @@ pgap2/lib/phylogeny.py
 pgap2/lib/pklcheck.py
 pgap2/lib/species.py
 pgap2/lib/strain.py
+pgap2/lib/synteny.py
 pgap2/lib/tree.py
+pgap2/models/__init__.py
+pgap2/models/baps.py
+pgap2/models/base.py
+pgap2/models/gene_annotation.py
+pgap2/models/gene_cluster.py
+pgap2/models/gene_neighborhood.py
+pgap2/models/pav.py
+pgap2/models/phylogeny.py
+pgap2/models/preprocess.py
+pgap2/models/project.py
+pgap2/models/project_file.py
+pgap2/models/statistics.py
+pgap2/models/strain.py
+pgap2/models/tajimas_d.py
 pgap2/postprocess/__init__.py
 pgap2/postprocess/baps.py
 pgap2/postprocess/graph.py
+pgap2/postprocess/neighborhood.py
 pgap2/postprocess/profile.py
 pgap2/postprocess/reference_layout.py
 pgap2/postprocess/singletree.py
@@ -53,14 +71,18 @@ pgap2/utils/draw.py
 pgap2/utils/gene_retriever.py
 pgap2/utils/generate_tree.py
 pgap2/utils/graph_utils.py
+pgap2/utils/neighborhood.py
 pgap2/utils/partition.py
 pgap2/utils/partition_add.py
 pgap2/utils/partition_merge.py
 pgap2/utils/partition_minus.py
 pgap2/utils/postprocess.py
 pgap2/utils/preprocess.py
+pgap2/utils/sqlite_builder.py
 pgap2/utils/supply.py
 pgap2/utils/tools.py
 tests/test_main.py
+tests/test_pangenome.py
 tests/test_post.py
-tests/test_prep.py
+tests/test_prep.py
+tests/test_synteny.py

{pgap2-2.0 → pgap2-2.3}/PGAP2.egg-info/requires.txt

@@ -14,3 +14,5 @@ ete3==3.1.3
 svgwrite==1.4.3
 miniFasta==3.0.2
 tajimas_d==2.0.2
+sqlalchemy>=2.0
+igraph

{pgap2-2.0/PGAP2.egg-info → pgap2-2.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: PGAP2
-Version: 2.0
+Version: 2.3
 Summary: PGAP2: a comprehensive pan-genome analysis pipeline for prokaryotic genomes
 Author-email: bucongfan <bucf@big.ac.cn>
 Maintainer-email: bucongfan <bucf@big.ac.cn>
@@ -28,6 +28,8 @@ Requires-Dist: ete3==3.1.3
 Requires-Dist: svgwrite==1.4.3
 Requires-Dist: miniFasta==3.0.2
 Requires-Dist: tajimas_d==2.0.2
+Requires-Dist: sqlalchemy>=2.0
+Requires-Dist: igraph
 Dynamic: license-file
 
 ![image](https://github.com/bucongfan/PGAP2/blob/main/pgap2.logo.png)

pgap2-2.3/pgap2/cli/neighborhood.py

@@ -0,0 +1,209 @@
+"""
+CLI definitions for the 'neigh' (conserved gene neighborhood) subcommand.
+
+This subcommand runs the full partition pipeline (identical to 'main')
+followed by a conserved gene neighborhood discovery step inspired by
+spacedust's agglomerative clustering algorithm.
+"""
+
+import os
+import argparse
+
+from argparse import ArgumentParser, _SubParsersAction
+
+from pgap2.utils.supply import sfw, tqdm_
+from pgap2.utils.supply import set_verbosity_level
+from pgap2.utils.graph_utils import check_min_falen, check_gcode
+
+
+def launch(args: argparse.Namespace):
+    from pgap2.utils.neighborhood import main
+    main(
+        indir=os.path.abspath(args.indir),
+        outdir=os.path.abspath(args.outdir),
+        falen=args.min_falen, threads=args.threads, evalue=args.evalue,
+        aligner=args.aligner, clust_method=args.clust_method,
+        orth_id=args.orth_id, para_id=args.para_id, dup_id=args.dup_id,
+        id_attr_key=args.id_attr_key, type_filter=args.type_filter,
+        coverage=0.98, fast_mode=args.fast_mode, hconf_thre=args.hconf_thre,
+        LD=args.LD, AS=args.AS, AL=args.AL, max_targets=args.max_targets,
+        context_similarity=0, flank=5,
+        accurate=args.accurate,
+        exhaust_orth=args.exhaust_orth,
+        gcode=args.gcode,
+        disable=args.disable, annot=args.annot, retrieve=args.retrieve,
+        radius=args.radius, sensitivity=args.sensitivity, ins=args.ins,
+        # neighborhood-specific parameters
+        max_gap=args.max_gap,
+        min_genes=args.min_genes,
+        pval_threshold=args.pval,
+        min_genomes=args.min_genomes,
+        resolution=args.resolution,
+        use_order=args.use_order,
+    )
+
+
+def neighborhood_portal(args):
+    set_verbosity_level(args.outdir, args.verbose,
+                        args.debug, 'neighborhood')
+
+    if args.clust_method == 'mmseqs2':
+        sfw.check_dependency("mmseqs2")
+    elif args.clust_method == 'cdhit':
+        sfw.check_dependency("cdhit")
+
+    sfw.check_dependency("mcl")
+    sfw.check_dependency("mcxdeblast")
+    if args.aligner == 'diamond':
+        sfw.check_dependency("diamond")
+    elif args.aligner == 'blastp':
+        sfw.check_dependency("blastp")
+        sfw.check_dependency("makeblastdb")
+
+    if args.retrieve:
+        sfw.check_dependency("miniprot")
+        sfw.check_dependency("seqtk")
+        tqdm_.set_total_step(14)  # 12 partition + 2 neighborhood steps
+    else:
+        tqdm_.set_total_step(9)   # 7 partition + 2 neighborhood steps
+
+    if args.annot:
+        sfw.check_dependency("prodigal")
+
+    launch(args)
+
+
+def neighborhood_cmd(subparser: _SubParsersAction):
+
+    subparser_neigh: ArgumentParser = subparser.add_parser(
+        'neigh',
+        help='Conserved gene neighborhood (synteny block) analysis. '
+             'Runs the full partition pipeline then discovers conserved '
+             'gene neighborhoods across genomes (spacedust-style).',
+        formatter_class=argparse.ArgumentDefaultsHelpFormatter)
+
+    # ---- Partition parameters (same as 'main') ----
+    subparser_neigh.add_argument(
+        '--indir', '-i', required=True,
+        help='Input directory containing genome files. Same prefix = same strain.')
+    subparser_neigh.add_argument(
+        '--outdir', '-o', required=False, default='.',
+        help='Output directory')
+    subparser_neigh.add_argument(
+        '--dup_id', required=False, type=float, default=0.99,
+        help='Maximum identity for most recent duplication events.')
+    subparser_neigh.add_argument(
+        '--orth_id', required=False, type=float, default=0.98,
+        help='Maximum identity between most similar pan-clusters.')
+    subparser_neigh.add_argument(
+        '--para_id', required=False, type=float, default=0.7,
+        help='Paralogous identity threshold.')
+    subparser_neigh.add_argument(
+        '--type-filter', required=False, type=str, default='CDS',
+        help='GFF feature type (3rd column) to include.')
+    subparser_neigh.add_argument(
+        '--id-attr-key', required=False, type=str, default='ID',
+        help='GFF attribute key for record ID.')
+    subparser_neigh.add_argument(
+        '--hconf_thre', required=False, type=float, default=1,
+        help='High confidence cluster threshold.')
+    subparser_neigh.add_argument(
+        '--exhaust_orth', '-e', required=False, action='store_true',
+        default=False, help='Exhaustive paralog splitting.')
+    subparser_neigh.add_argument(
+        '--sensitivity', '-s', required=False, type=str, default='strict',
+        choices=('soft', 'moderate', 'strict'),
+        help='Node connectivity sensitivity.')
+    subparser_neigh.add_argument(
+        '--ins', '-n', required=False, action='store_true', default=False,
+        help='Ignore insertion sequence influence.')
+    subparser_neigh.add_argument(
+        '--fast', '-f', dest='fast_mode', required=False,
+        action='store_true', default=False,
+        help='Fast mode (skip fine feature analysis).')
+    subparser_neigh.add_argument(
+        '--accurate', '-a', required=False, action='store_true',
+        default=False, help='Bidirection check for paralog partition.')
+    subparser_neigh.add_argument(
+        '--retrieve', '-r', required=False, action='store_true',
+        default=False, help='Retrieve genes lost during annotation.')
+    subparser_neigh.add_argument(
+        '--threads', '-t', required=False, default=1, type=int,
+        help='Number of threads.')
+    subparser_neigh.add_argument(
+        '--max_targets', '-k', required=False, type=int, default=2000,
+        help='Maximum alignment targets per query.')
+    subparser_neigh.add_argument(
+        '--LD', required=False, type=float, default=0.6,
+        help='Minimum gene length difference proportion.')
+    subparser_neigh.add_argument(
+        '--AS', required=False, type=float, default=0.6,
+        help='Coverage for shorter sequence.')
+    subparser_neigh.add_argument(
+        '--AL', required=False, type=float, default=0.6,
+        help='Coverage for longer sequence.')
+    subparser_neigh.add_argument(
+        '--evalue', required=False, type=float, default=1E-5,
+        help='Aligner E-value threshold.')
+    subparser_neigh.add_argument(
+        '--aligner', required=False, type=str, default='diamond',
+        choices=('diamond', 'blastp'), help='Pairwise aligner.')
+    subparser_neigh.add_argument(
+        '--clust_method', required=False, type=str, default='mmseqs2',
+        choices=('cdhit', 'mmseqs2'), help='Gene clustering method.')
+    subparser_neigh.add_argument(
+        '--radius', required=False, type=int, default=3,
+        help='Search region radius.')
+    subparser_neigh.add_argument(
+        '--min_falen', '-m', required=False, type=check_min_falen,
+        default=20, help='Minimum protein length (at least 11).')
+    subparser_neigh.add_argument(
+        '--gcode', required=False, type=check_gcode, default=11,
+        help='Genetic code.')
+    subparser_neigh.add_argument(
+        '--annot', required=False, action='store_true', default=False,
+        help='Re-annotate genomes with Prodigal.')
+
+    # ---- Neighborhood-specific parameters ----
+    neigh_group = subparser_neigh.add_argument_group(
+        'Neighborhood parameters',
+        'Parameters specific to conserved gene neighborhood detection.')
+    neigh_group.add_argument(
+        '--max_gap', required=False, type=int, default=3,
+        help='Maximum gene gap allowed between hits for merging into a '
+             'syntenic block. Larger values allow more intervening genes.')
+    neigh_group.add_argument(
+        '-mg','--min_genes', required=False, type=int, default=2,
+        help='Minimum number of genes to define a syntenic block.')
+    neigh_group.add_argument(
+        '--pval', required=False, type=float, default=0.01,
+        help='P-value threshold for gene-order filtering (only used '
+             'when --use_order is enabled).')
+    neigh_group.add_argument(
+        '--use_order', required=False, action='store_true', default=False,
+        help='Enable gene-order conservation filtering. When set, blocks '
+             'are filtered by P_order <= --pval, edge weights use '
+             '-log10(P_order), and neighborhood P-values are computed '
+             'via Fisher\'s combined test. Default: off (accept blocks '
+             'purely based on --min_genes co-localization).')
+    neigh_group.add_argument(
+        '-mgs','--min_genomes', required=False, type=int, default=2,
+        help='Minimum number of genomes a neighborhood must span to be reported.')
+    neigh_group.add_argument(
+        '-re','--resolution', required=False, type=float, default=1.0,
+        help='Louvain resolution parameter. Higher values produce smaller, '
+             'more focused neighborhoods; lower values produce larger, '
+             'more inclusive ones. Try 2.0-10.0 if neighborhoods are too large.')
+
+    # ---- Common flags ----
+    subparser_neigh.add_argument(
+        '--verbose', '-v', required=False, action='store_true',
+        default=False, help='Verbose output')
+    subparser_neigh.add_argument(
+        '--debug', '-D', required=False, action='store_true',
+        default=False, help='Debug mode (very verbose)')
+    subparser_neigh.add_argument(
+        '--disable', required=False, action='store_true',
+        default=False, help='Disable progress bar')
+
+    subparser_neigh.set_defaults(func=neighborhood_portal)

{pgap2-2.0 → pgap2-2.3}/pgap2/cli/postprocess.py

@@ -7,6 +7,7 @@ from pgap2.postprocess.baps import baps_cmd
 from pgap2.postprocess.singletree import singletree_cmd
 from pgap2.postprocess.tajimas_d import tajimas_d_cmd
 from pgap2.postprocess.graph import graph_cmd
+from pgap2.postprocess.neighborhood import neigh_cmd
 from argparse import ArgumentParser, _SubParsersAction
 import argparse
 
@@ -21,3 +22,4 @@ def postprocess_cmd(subparser: _SubParsersAction):
     baps_cmd(sub_subparsers)
     tajimas_d_cmd(sub_subparsers)
     graph_cmd(sub_subparsers)
+    neigh_cmd(sub_subparsers)

pgap2-2.3/pgap2/cli/sqlite_cmd.py

@@ -0,0 +1,84 @@
+"""CLI definitions for the ``pgap2 sqlite`` subcommand.
+
+Generates a self-contained SQLite database from PGAP2 output files.
+
+Usage::
+
+    pgap2 sqlite -i /path/to/output
+    pgap2 sqlite -i /path/to/output -o /path/to/output/pgap2.db
+    pgap2 sqlite -i /path/to/output --name "My Project" --species "E. coli"
+"""
+
+import os
+import argparse
+from argparse import ArgumentParser, _SubParsersAction
+from pathlib import Path
+
+
+def launch(args: argparse.Namespace):
+    from pgap2.utils.sqlite_builder import build_sqlite, parse_readme
+
+    indir = os.path.abspath(args.indir)
+
+    # Resolve output path: default is <indir>/pgap2.db
+    if args.output:
+        output_path = os.path.abspath(args.output)
+    else:
+        output_path = os.path.join(indir, "pgap2.db")
+
+    # Try to derive project name / species from README.md if not provided
+    project_name = args.name
+    species = args.species
+
+    readme_path = Path(indir) / "README.md"
+    if readme_path.exists():
+        readme_info = parse_readme(readme_path)
+        if not project_name:
+            project_name = readme_info.get("project_name", "")
+        if not species:
+            species = readme_info.get("species", "")
+
+    if not project_name:
+        project_name = Path(indir).name
+    if not species:
+        species = "Unknown"
+
+    info = build_sqlite(
+        data_dir=indir,
+        output_path=output_path,
+        project_name=project_name,
+        species=species,
+        description=args.description or "",
+    )
+
+    return 0
+
+
+def sqlite_portal(args):
+    """Entry point called by argparse (dependency checks + launch)."""
+    launch(args)
+
+
+def sqlite_cmd(subparser: _SubParsersAction):
+    """Register the ``sqlite`` subcommand."""
+    sp: ArgumentParser = subparser.add_parser(
+        "sqlite",
+        help="Generate a SQLite database from PGAP2 output files",
+        formatter_class=argparse.ArgumentDefaultsHelpFormatter,
+    )
+    sp.add_argument(
+        "--indir", "-i", required=True,
+        help="PGAP2 output directory (contains .pav, .detail.tsv, etc.)")
+    sp.add_argument(
+        "--output", "-o", required=False, default=None,
+        help="Output SQLite file path (default: <indir>/pgap2.db)")
+    sp.add_argument(
+        "--name", "-n", required=False, default=None,
+        help="Project name (auto-detected from README.md if present)")
+    sp.add_argument(
+        "--species", "-s", required=False, default=None,
+        help="Species name (auto-detected from README.md if present)")
+    sp.add_argument(
+        "--description", "-d", required=False, default="",
+        help="Optional project description")
+    sp.set_defaults(func=sqlite_portal)

{pgap2-2.0 → pgap2-2.3}/pgap2/lib/pangenome.py

@@ -272,8 +272,7 @@ class Pangenome():
             bar = tqdm(total=self.total_gene_num,
                        unit=' Gene', desc=tqdm_.step(6), disable=self.disable_tqdm)
             loaded_contig_name = set()
-            if retrieve:
-                coord_pat = re.compile(r'\[(\d+):\d+\]\(([+-])\)')
+            coord_pat = re.compile(r'\[(\d+):\d+\]\(([+-])\)')
 
             with open(flat_annot_file) as fh:
                 for line in fh:
@@ -283,28 +282,29 @@ class Pangenome():
                     bar.update()
                     gene_index = lines[0]
                     contig_name = lines[2]
-                    if contig_name not in loaded_contig_name:
-                        loaded_contig_name.add(contig_name)
-                        self.annot_contig_map.update(
-                            {':'.join(gene_index.split(':')[:2]): contig_name})
+                    contig_key = ':'.join(gene_index.split(':')[:2])
+                    if contig_key not in loaded_contig_name:
+                        #may duplicate when use contig name as the global var, chaning to genome_id:contig_id
+                        loaded_contig_name.add(contig_key)
+                        self.annot_contig_map[contig_key] = contig_name
                     gene_len = int(lines[4])
                     gene_id = lines[5]
                     gene_name = lines[6]
                     gene_product = lines[7]
+                    strand = ''
+                    m = coord_pat.search(lines[3])
+                    if m:
+                        start = int(m.group(1))
+                        strand = m.group(2)        # '+' or '-'
                     if retrieve:
-                        # start = int(re.search(r'\[(\d+):', lines[3]).group(1))
-                        # strand = re.search(r'\[(\+|-)\d+:\d+\]', lines[3]).group(1)
-                        m = coord_pat.search(lines[3])
-                        if m:
-                            start = int(m.group(1))   # 起始坐标
-                            strand = m.group(2)        # '+' 或 '-'
                         contig = ":".join(lines[0].split(":")[:2])
                         if strand == '+':
                             self.gene_rank[contig][0].append(start)
                         else:
                             self.gene_rank[contig][1].append(start)
                     self.annot.update({gene_index: {'len': gene_len, 'id': gene_id,
-                                                    'name': gene_name, 'product': gene_product}})
+                                                    'name': gene_name, 'product': gene_product,
+                                                    'strand': strand}})
             bar.close()
         else:
             logger.error(f'Cannot find the annot file {flat_annot_file}')

{pgap2-2.0 → pgap2-2.3}/pgap2/lib/phylogeny.py

@@ -45,7 +45,7 @@ def set_logger(logger_):
 
 
 class Phylogeny():
-    def __init__(self, basic, outdir, threads, disable, msa_method, tree_method, fastbaps_levels=None, fastbaps_prior=None, add_paras=[]) -> None:
+    def __init__(self, basic, outdir, threads, disable, msa_method, tree_method, fastbaps_levels=None, fastbaps_prior=None, add_paras=[], no_partition=False) -> None:
         self.basic: Basic = basic
         self.outdir = outdir
         self.threads = threads
@@ -54,6 +54,7 @@ class Phylogeny():
         self.tree_method = tree_method
         self.fastbaps_levels = fastbaps_levels
         self.fastbaps_prior = fastbaps_prior
+        self.no_partition = no_partition
         self.add_paras_dict = {2: '', 4: '',
                                6: '', 7: '', 8: '', 9: '', 10: ''}
         self.load_additional_paras(add_paras)
@@ -359,6 +360,7 @@ class Phylogeny():
         concatenated_gb.annotations['molecule_type'] = 'DNA'
 
         position_offset = 0  # Track the concatenated sequence length
+        partition_entries = []  # RAxML-style partition definitions
 
         real_involved_clusters = {
             strain: 0 for strain in self.basic.strain_dict}
@@ -400,8 +402,23 @@ class Phylogeny():
                                  type="region",
                                  qualifiers={"note": core_cluster_path, "locus_tag": core_cluster})
             concatenated_gb.features.append(feature)
+            # Record partition range (1-based) for RAxML/IQ-TREE partition model
+            partition_entries.append(
+                f'GTR+G, {core_cluster} = {position_offset + 1}-{position_offset + example_length}')
             position_offset += example_length  # Update position for the next cluster
 
+        # Save partition file (RAxML-style, also accepted by IQ-TREE with -p)
+        if not self.no_partition:
+            partition_path = os.path.join(
+                cat_align_outdir, 'core_gene_alignment.partitions')
+            with open(partition_path, 'w') as fh:
+                fh.write('\n'.join(partition_entries) + '\n')
+            logger.info(
+                f'Partition file: {partition_path} ({len(partition_entries)} partitions)')
+        else:
+            partition_path = None
+            logger.info('Partition file generation skipped (--no-partition)')
+
         # Save the concatenated record to a GenBank file
         SeqIO.write(concatenated_gb, os.path.join(
             cat_align_outdir, 'core_gene_alignment.gb'), 'genbank')
@@ -424,6 +441,9 @@ class Phylogeny():
         self.results_file.append(f'{cat_align_outdir}/core_gene_alignment.aln')
         self.results_file.append(f'{cat_align_outdir}/core_gene_alignment.gb')
         self.results_file.append(f'{cat_align_outdir}/core_gene_alignment.txt')
+        if partition_path is not None:
+            self.results_file.append(partition_path)
+            self.partition_file = partition_path
         return f'{cat_align_outdir}/core_gene_alignment.aln'
 
     def construct_tree(self, wd):
@@ -433,18 +453,29 @@ class Phylogeny():
         os.makedirs(tree_outdir, exist_ok=True)
         tree_path = os.path.join(tree_outdir, 'core_gene_alignment')
         add_paras = self.add_paras_dict[6]
+        extra = (add_paras or "").strip()
+        model_arg = 'GTR+G'
+
         if self.tree_method == 'raxml':
-            cmd = f'{sfw.raxml} --msa {self.first_aln} --model GTR+G --prefix {tree_path} --all --bs-trees 100 --redo {add_paras}'
-            best_tree = os.path.join(
-                tree_outdir, 'core_gene_alignment'+'.raxml.bestTree')
+            cmd = (
+                f'{sfw.raxml} --msa {self.first_aln} --model {model_arg} '
+                f'--prefix {tree_path} --all --bs-trees 100 --redo {extra}'
+            ).strip()
+            best_tree = f'{tree_path}.raxml.support'
+
         elif self.tree_method == 'fasttree':
-            cmd = f'{sfw.fasttree} -gtr -gamma -nt -out {tree_path}.treefile {self.first_aln} {add_paras} 2> {tree_path}.log'
-            best_tree = os.path.join(
-                tree_outdir, 'core_gene_alignment'+'.treefile')
+            cmd = (
+                f'{sfw.fasttree} -nt -gtr -gamma {extra} '
+                f'-log {tree_path}.log {self.first_aln} > {tree_path}.treefile'
+            ).strip()
+            best_tree = f'{tree_path}.treefile'
+
         elif self.tree_method == 'iqtree':
-            cmd = f'{sfw.iqtree} -s {self.first_aln} -m GTR+G -pre {tree_path} -bb 1000 -redo -nt AUTO {add_paras}'
-            best_tree = os.path.join(
-                tree_outdir, 'core_gene_alignment'+'.treefile')
+            cmd = (
+                f'{sfw.iqtree} -s {self.first_aln} -m {model_arg} '
+                f'-pre {tree_path} -bb 1000 -bnni -redo -nt AUTO {extra}'
+            ).strip()
+            best_tree = f'{tree_path}.treefile'
 
         self.results_file.append(best_tree)
 
@@ -529,19 +560,40 @@ class Phylogeny():
         os.makedirs(tree_outdir, exist_ok=True)
         tree_path = os.path.join(tree_outdir, 'reconstructed_tree')
         add_paras = self.add_paras_dict[9]
+        extra = (add_paras or "").strip()
+
+        # Use partition model when partition file is available
+        has_partition = hasattr(
+            self, 'partition_file') and os.path.exists(self.partition_file)
 
         if self.tree_method == 'raxml':
-            cmd = f'{sfw.raxml} --msa {self.maskrc_aln} --model GTR+G --prefix {tree_path} --all --bs-trees 100 --redo {add_paras}'
-            best_tree = os.path.join(
-                tree_outdir, 'reconstructed_tree'+'.raxml.bestTree')
+            model_arg = self.partition_file if has_partition else 'GTR+G'
+            cmd = (
+                f'{sfw.raxml} --msa {self.maskrc_aln} --model {model_arg} '
+                f'--prefix {tree_path} --all --bs-trees 100 --redo {extra}'
+            ).strip()
+            best_tree = f'{tree_path}.raxml.support'   # 带 bootstrap 支持值的最终树
+
         elif self.tree_method == 'fasttree':
-            cmd = f'{sfw.fasttree} -gtr -gamma -nt -out {tree_path}.treefile {self.maskrc_aln} {add_paras} 2> {tree_path}.log'
-            best_tree = os.path.join(
-                tree_outdir, 'reconstructed_tree'+'.treefile')
+            # FastTree does not support partition models
+            cmd = (
+                f'{sfw.fasttree} -nt -gtr -gamma {extra} '
+                f'-log {tree_path}.log {self.maskrc_aln} > {tree_path}.treefile'
+            ).strip()
+            best_tree = f'{tree_path}.treefile'
+
         elif self.tree_method == 'iqtree':
-            cmd = f'{sfw.iqtree} -s {self.maskrc_aln} -m GTR+G -pre {tree_path} -bb 1000 -redo -nt AUTO {add_paras}'
-            best_tree = os.path.join(
-                tree_outdir, 'reconstructed_tree'+'.treefile')
+            if has_partition:
+                cmd = (
+                    f'{sfw.iqtree} -s {self.maskrc_aln} -p {self.partition_file} '
+                    f'-pre {tree_path} -m MFP+MERGE -bb 1000 -bnni -redo -nt AUTO {extra}'
+                ).strip()
+            else:
+                cmd = (
+                    f'{sfw.iqtree} -s {self.maskrc_aln} -m GTR+G '
+                    f'-pre {tree_path} -bb 1000 -bnni -redo -nt AUTO {extra}'
+                ).strip()
+            best_tree = f'{tree_path}.treefile'
 
         self.results_file.append(best_tree)
         finished_status = self.check_before_run(cmd, tree_outdir)