modelcraft 5.0.2__py3-none-any.whl → 6.0.0__py3-none-any.whl

modelcraft/coot/prune.py

@@ -1,1085 +0,0 @@
-import math
-
-# CONSTS
-
-atomic_numbers = {
-    "H": 1,
-    "HE": 2,
-    "LI": 3,
-    "BE": 4,
-    "B": 5,
-    "C": 6,
-    "N": 7,
-    "O": 8,
-    "F": 9,
-    "NE": 10,
-    "NA": 11,
-    "MG": 12,
-    "AL": 13,
-    "SI": 14,
-    "P": 15,
-    "S": 16,
-    "CL": 17,
-    "AR": 18,
-    "K": 19,
-    "CA": 20,
-    "SC": 21,
-    "TI": 22,
-    "V": 23,
-    "CR": 24,
-    "MN": 25,
-    "FE": 26,
-    "CO": 27,
-    "NI": 28,
-    "CU": 29,
-    "ZN": 30,
-    "GA": 31,
-    "GE": 32,
-    "AS": 33,
-    "SE": 34,
-    "BR": 35,
-    "KR": 36,
-    "RB": 37,
-    "SR": 38,
-    "Y": 39,
-    "ZR": 40,
-    "NB": 41,
-    "MO": 42,
-    "TC": 43,
-    "RU": 44,
-    "RH": 45,
-    "PD": 46,
-    "AG": 47,
-    "CD": 48,
-    "IN": 49,
-    "SN": 50,
-    "SB": 51,
-    "TE": 52,
-    "I": 53,
-    "XE": 54,
-    "CS": 55,
-    "BA": 56,
-    "LA": 57,
-    "CE": 58,
-    "PR": 59,
-    "ND": 60,
-    "PM": 61,
-    "SM": 62,
-    "EU": 63,
-    "GD": 64,
-    "TB": 65,
-    "DY": 66,
-    "HO": 67,
-    "ER": 68,
-    "TM": 69,
-    "YB": 70,
-    "LU": 71,
-    "HF": 72,
-    "TA": 73,
-    "W": 74,
-    "RE": 75,
-    "OS": 76,
-    "IR": 77,
-    "PT": 78,
-    "AU": 79,
-    "HG": 80,
-    "TL": 81,
-    "PB": 82,
-    "BI": 83,
-    "PO": 84,
-    "AT": 85,
-    "RN": 86,
-    "FR": 87,
-    "RA": 88,
-    "AC": 89,
-    "TH": 90,
-    "PA": 91,
-    "U": 92,
-    "NP": 93,
-    "PU": 94,
-    "AM": 95,
-    "CM": 96,
-    "BK": 97,
-    "CF": 98,
-    "ES": 99,
-    "FM": 100,
-    "MD": 101,
-    "NO": 102,
-    "LR": 103,
-    "RF": 104,
-    "DB": 105,
-    "SG": 106,
-    "BH": 107,
-    "HS": 108,
-    "MT": 109,
-    "DS": 110,
-    "RG": 111,
-    "CN": 112,
-    "NH": 113,
-    "FL": 114,
-    "MC": 115,
-    "LV": 116,
-    "TS": 117,
-    "OG": 118,
-}
-
-protein_residues = {
-    "ALA",
-    "ARG",
-    "ASN",
-    "ASP",
-    "CYS",
-    "GLN",
-    "GLU",
-    "GLY",
-    "HIS",
-    "ILE",
-    "LEU",
-    "LYS",
-    "MET",
-    "MSE",
-    "PHE",
-    "PRO",
-    "SER",
-    "THR",
-    "TRP",
-    "TYR",
-    "UNK",
-    "VAL",
-}
-
-main_chain_atoms = {" N  ", " CA ", " C  ", " O  ", " CB "}
-
-bonded = {
-    " C  ": {" CA ", " O  ", " OXT"},
-    " CA ": {" C  ", " CB ", " N  "},
-    " CB ": {" CA ", " CG ", " CG1", " CG2", " OG ", " OG1", " SG "},
-    " CD ": {" CE ", " CG ", " NE ", " NE2", " OE1", " OE2"},
-    " CD1": {" CE1", " CG ", " CG1", " NE1"},
-    " CD2": {" CE2", " CE3", " CG ", " NE2"},
-    " CE ": {" CD ", " NZ ", " SD ", "SE  "},
-    " CE1": {" CD1", " CZ ", " ND1", " NE2"},
-    " CE2": {" CD2", " CZ ", " CZ2", " NE1"},
-    " CE3": {" CD2", " CZ3"},
-    " CG ": {" CB ", " CD ", " CD1", " CD2", " ND1", " ND2", " OD1", " OD2", " SD ", "SE  ",},
-    " CG1": {" CB ", " CD1"},
-    " CG2": {" CB "},
-    " CH2": {" CZ2", " CZ3"},
-    " CZ ": {" CE1", " CE2", " NE ", " NH1", " NH2", " OH "},
-    " CZ2": {" CE2", " CH2"},
-    " CZ3": {" CE3", " CH2"},
-    " N  ": {" CA "},
-    " ND1": {" CE1", " CG "},
-    " ND2": {" CG "},
-    " NE ": {" CD ", " CZ "},
-    " NE1": {" CD1", " CE2"},
-    " NE2": {" CD ", " CD2", " CE1"},
-    " NH1": {" CZ "},
-    " NH2": {" CZ "},
-    " NZ ": {" CE "},
-    " O  ": {" C  "},
-    " OD1": {" CG "},
-    " OD2": {" CG "},
-    " OE1": {" CD "},
-    " OE2": {" CD "},
-    " OG ": {" CB "},
-    " OG1": {" CB "},
-    " OH ": {" CZ "},
-    " OXT": {" C  "},
-    " SD ": {" CE ", " CG "},
-    " SG ": {" CB "},
-    "SE  ": {" CE ", " CG "},
-}
-
-# UTILS
-
-
-def mean(values):
-    return float(sum(values)) / len(values)
-
-
-def median(values):
-    n = len(values)
-    if n < 1:
-        return None
-    i = n // 2
-    if n % 2 == 1:
-        return sorted(values)[i]
-    return sum(sorted(values)[i - 1 : i + 1]) / 2.0
-
-
-def median_absolute_deviation(values):
-    median_value = median(values)
-    abs_deviations = [abs(value - median_value) for value in values]
-    return median(abs_deviations)
-
-
-def dot(xyz1, xyz2):
-    return xyz1[0] * xyz2[0] + xyz1[1] * xyz2[1] + xyz1[2] * xyz2[2]
-
-
-def cross(xyz1, xyz2):
-    return [
-        xyz1[1] * xyz2[2] - xyz1[2] * xyz2[1],
-        xyz1[2] * xyz2[0] - xyz1[0] * xyz2[2],
-        xyz1[0] * xyz2[1] - xyz1[1] * xyz2[0],
-    ]
-
-
-def magnitude(xyz):
-    return (xyz[0] ** 2 + xyz[1] ** 2 + xyz[2] ** 2) ** 0.5
-
-
-def unit(xyz):
-    length = magnitude(xyz)
-    return [xyz[0] / length, xyz[1] / length, xyz[2] / length]
-
-
-def subtract(xyz1, xyz2):
-    return [xyz1[0] - xyz2[0], xyz1[1] - xyz2[1], xyz1[2] - xyz2[2]]
-
-
-def distance(xyz1, xyz2):
-    v = subtract(xyz1, xyz2)
-    return magnitude(v)
-
-
-def angle(xyz1, xyz2, xyz3):
-    v1 = subtract(xyz2, xyz1)
-    v2 = subtract(xyz2, xyz3)
-    angle = math.acos(dot(v1, v2) / (magnitude(v1) * magnitude(v2)))
-    return math.degrees(angle)
-
-
-def torsion(xyz1, xyz2, xyz3, xyz4):
-    b1 = subtract(xyz2, xyz1)
-    b2 = subtract(xyz3, xyz2)
-    b3 = subtract(xyz4, xyz3)
-    n1 = cross(b1, b2)
-    n2 = cross(b2, b3)
-    m1 = cross(n1, n2)
-    y = dot(m1, unit(b2))
-    x = dot(n1, n2)
-    angle = math.degrees(math.atan2(y, x))
-    if angle > 180:
-        angle -= 360
-    return angle
-
-
-def halfway(xyz1, xyz2):
-    x = (xyz1[0] + xyz2[0]) / 2
-    y = (xyz1[1] + xyz2[1]) / 2
-    z = (xyz1[2] + xyz2[2]) / 2
-    return [x, y, z]
-
-
-def attached_atoms(atom, residue):
-    if atom.name == " N  " and residue.prev is not None:
-        yield residue.prev.atoms[" C  "]
-    if atom.name == " C  " and residue.next is not None:
-        yield residue.next.atoms[" N  "]
-    for other in bonded[atom.name]:
-        if other in residue.atoms:
-            yield residue.atoms[other]
-
-
-# MACHINE LEARNING
-
-training_data = {
-    "main": {
-        "medians": [
-            0.8093730211257935,
-            0.0,
-            0.3797053494361412,
-            0.2948472222222224,
-            1.1603445955372607,
-            0.07967734307025365,
-            -1.04405097291337,
-            -0.7666229846405788,
-            0.09301174988214152,
-            0.2074185971626753,
-            1.9998366832733154,
-            7.473087887125843,
-        ],
-        "scaler.mean_": [
-            0.7635879979927991,
-            0.0119850162929945,
-            0.47070085149410484,
-            0.8779728019083975,
-            1.4010333429538593,
-            -0.07809440994354622,
-            -1.1634877907354524,
-            -0.9231595065864178,
-            0.06630610233126723,
-            0.7376180259790496,
-            1.9333955416970983,
-            9.00267277208097,
-        ],
-        "scaler.scale_": [
-            0.15510631635324296,
-            0.10881808524983969,
-            0.49424003745942746,
-            2.1942551160797694,
-            1.298509344613297,
-            0.8836945495954351,
-            0.9345174344061477,
-            0.9874691652931548,
-            1.2037765763286472,
-            1.1729051296356117,
-            0.42936735360011963,
-            6.84475636137809,
-        ],
-        "regressor.coefs_": [
-            [
-                [
-                    -0.13040818548223823,
-                    -0.22295218610758016,
-                    -0.16258080220962234,
-                    0.5988690897292985,
-                    0.29886699802544087,
-                    -0.2138536188789283,
-                    0.17640972642170277,
-                    0.8323631686040229,
-                    0.8080535193667213,
-                    -0.64199269078319,
-                ],
-                [
-                    0.3371557445052779,
-                    -0.021990355769565446,
-                    0.3393433644319566,
-                    0.7031031102737114,
-                    -1.5952677749631456,
-                    -0.6291226390140611,
-                    -0.16470505242375907,
-                    0.8959152619464158,
-                    0.8165087694148113,
-                    1.3192349998470725,
-                ],
-                [
-                    -0.0019874340519655917,
-                    0.04016420712930254,
-                    -0.0016519722530829112,
-                    -0.05753723016959708,
-                    -1.0970634376604886,
-                    0.13781915021293673,
-                    -0.043726103959200804,
-                    0.010119141446418834,
-                    -0.09415567058366356,
-                    0.10675615201134665,
-                ],
-                [
-                    -0.15553908141483186,
-                    0.7815309718136879,
-                    -0.58301702342861,
-                    -0.3074194496122189,
-                    -0.708058122322275,
-                    0.01629181598284882,
-                    -0.17713645375637574,
-                    -0.5780953291256867,
-                    -0.08330249685202995,
-                    -0.046797521299526336,
-                ],
-                [
-                    -0.16382881259387705,
-                    0.1590019158219002,
-                    -0.15271138566338957,
-                    0.004631275745636504,
-                    0.14085902736133887,
-                    0.044487235857714004,
-                    -0.22845244910340506,
-                    0.03361790942261084,
-                    -0.09431223857684323,
-                    0.1894309630932748,
-                ],
-                [
-                    0.4600605680220205,
-                    1.0475854981548196,
-                    0.28528470975660497,
-                    -0.07122257966372195,
-                    -0.17193073732931968,
-                    -0.563441753208365,
-                    -0.6552700463525175,
-                    -0.08483854601112313,
-                    -0.17645291495517423,
-                    -0.44659761042818563,
-                ],
-                [
-                    -0.32437585347959863,
-                    -0.40621672156363603,
-                    -0.22786021141673057,
-                    -0.006696637890890252,
-                    0.5191491920623607,
-                    0.16595743564965426,
-                    0.27911102055393494,
-                    0.2640726988201516,
-                    -0.0016347547075830314,
-                    0.24770754766817038,
-                ],
-                [
-                    0.3728665898401538,
-                    0.07700714533797552,
-                    0.2311708633268853,
-                    -0.1401187037969537,
-                    0.30234045086128297,
-                    -0.24458025060957192,
-                    0.4302767405048504,
-                    -0.3703955591237981,
-                    0.3484976393563264,
-                    -0.5594717372377868,
-                ],
-                [
-                    0.00937493962029786,
-                    -0.04694394839682417,
-                    -0.028893614058871878,
-                    0.27141499045973605,
-                    -0.07941767649526653,
-                    -0.03911716545463574,
-                    -0.12103549705527984,
-                    0.2148437699515711,
-                    0.08720611510332744,
-                    -0.13302591593197644,
-                ],
-                [
-                    0.16013205898626096,
-                    0.35382721838123166,
-                    0.05844089574619886,
-                    -0.2090590614354888,
-                    0.06453171825149609,
-                    -1.172220172769279,
-                    -0.14500654761975515,
-                    -0.35695026385344086,
-                    0.7073129963007073,
-                    -1.1972037303520695,
-                ],
-                [
-                    0.12999852919812566,
-                    -0.37003217528080845,
-                    0.3596837609653873,
-                    0.30447584223950475,
-                    -0.5844046567044845,
-                    1.099321801469266,
-                    0.28973479295322985,
-                    0.6123382299120337,
-                    -0.8254996440849733,
-                    0.3860328709153097,
-                ],
-                [
-                    -0.013864646575082658,
-                    0.33883378654775376,
-                    -0.0432656081821551,
-                    -0.03831855907564628,
-                    -0.17009170626798892,
-                    0.09755681267845129,
-                    -0.14149706478955423,
-                    0.039883119397616046,
-                    -0.11814265819588236,
-                    0.13867097327343803,
-                ],
-            ],
-            [
-                [0.4113990249648804],
-                [-0.1635521743014414],
-                [-0.3233742954955824],
-                [-0.2832881263154175],
-                [0.10184600411275878],
-                [0.26297418395412697],
-                [-0.38667783069289213],
-                [0.18806411489930078],
-                [0.2650096089001783],
-                [-0.3432819700082653],
-            ],
-        ],
-        "regressor.intercepts_": [
-            [
-                0.05426873271484922,
-                -0.17185888723743994,
-                0.49670571655441914,
-                -0.26895026535845296,
-                1.2489563308304976,
-                -0.21430979896458663,
-                -1.0229946977640407,
-                0.43986781805151526,
-                0.32517336900734384,
-                -1.1981611190470425,
-            ],
-            [0.2009473880274795],
-        ],
-    },
-    "side": {
-        "medians": [
-            0.7128432393074036,
-            0.3930308684784101,
-            0.12138492565055738,
-            0.7697625388342151,
-            -0.01155913420195338,
-            -0.7577415671212863,
-            -0.5053093547058974,
-            1.9498454332351685,
-            36.654449462890625,
-        ],
-        "scaler.mean_": [
-            0.6699939922392967,
-            0.4861278093279958,
-            0.5598238115764408,
-            1.0098633759640239,
-            -0.09092168866565035,
-            -0.8734938716432699,
-            -0.5939212054114583,
-            1.903840522636135,
-            41.034051026184834,
-        ],
-        "scaler.scale_": [
-            0.1922931323119439,
-            0.5148420550064197,
-            1.547688727501242,
-            1.6216833496523257,
-            0.8535355005798227,
-            1.063733997556021,
-            0.946219729526465,
-            0.42543155678705524,
-            29.99679803225501,
-        ],
-        "regressor.coefs_": [
-            [
-                [
-                    0.056302872646141214,
-                    -0.42725673599695374,
-                    0.23328615715807235,
-                    -0.25692317829492684,
-                    0.116828236758043,
-                    1.2384242768052554,
-                    0.6748323891517939,
-                    0.14291234817220463,
-                    0.21380028409294824,
-                    0.07679070847654368,
-                ],
-                [
-                    -0.46153688659142184,
-                    -0.004544981320490466,
-                    -0.2964031182985578,
-                    0.09456605767558321,
-                    -0.06435341411291445,
-                    0.12673028609583875,
-                    -0.2719367005541086,
-                    0.29679974643773216,
-                    0.11486700383265894,
-                    0.23185475607178951,
-                ],
-                [
-                    0.33853298205315957,
-                    0.9079813683093967,
-                    0.09346553254580311,
-                    -0.025784490716986762,
-                    -0.3475803251546684,
-                    -0.04639625178653906,
-                    -0.500881871296081,
-                    0.7809152729815941,
-                    0.406203973798232,
-                    -0.3284891104603179,
-                ],
-                [
-                    -0.2088543439928225,
-                    0.47556925150407825,
-                    -0.19039046477157073,
-                    0.20421072806892146,
-                    -0.23471331429141476,
-                    -0.08725373460798075,
-                    -0.0811783537165059,
-                    0.4593053196175982,
-                    0.4351302763377739,
-                    0.11837686827471676,
-                ],
-                [
-                    -0.32138852798617806,
-                    0.11322678209797299,
-                    -0.07514559826050492,
-                    0.13497722508180546,
-                    0.2308583983764714,
-                    0.11180021860598335,
-                    -0.6957881978609368,
-                    0.0005190646203973872,
-                    -0.2880602141275108,
-                    -0.13675732785562053,
-                ],
-                [
-                    0.06423107339757392,
-                    0.08120627709915425,
-                    0.29375027114694363,
-                    -0.15903515362475384,
-                    -0.038018112519918414,
-                    0.2893167384125542,
-                    0.14378196316887357,
-                    0.04488068580666965,
-                    0.1342358964548304,
-                    0.00855401074520736,
-                ],
-                [
-                    -0.06837070927792858,
-                    -0.45192674269454314,
-                    0.059932235100582196,
-                    -0.15614805983481528,
-                    -0.3859760254395101,
-                    -0.3429573286705649,
-                    0.6838364312078055,
-                    -0.347976240289938,
-                    0.049432609557722565,
-                    -0.28055309729723454,
-                ],
-                [
-                    0.6431046107009092,
-                    0.49740363903294654,
-                    0.2748436188374371,
-                    -0.18326862373951083,
-                    -0.2069737876487165,
-                    -0.2999102167712396,
-                    -0.1590891264716532,
-                    0.13233313109972625,
-                    -0.030493374720953297,
-                    -0.324131016869104,
-                ],
-                [
-                    -0.44478515534248564,
-                    -0.22066143889313278,
-                    -0.31968023354915465,
-                    0.025341827810811048,
-                    -0.08457818649830318,
-                    0.16910787319016476,
-                    0.38203483276154415,
-                    -0.10511631297976766,
-                    0.08750071031831075,
-                    0.2747264545695668,
-                ],
-            ],
-            [
-                [0.10508025998686506],
-                [-0.2691991244124654],
-                [0.21154928243402793],
-                [0.6059699335423194],
-                [-0.3520197843417094],
-                [0.1473370411925756],
-                [0.3586311702546054],
-                [0.1428546809034874],
-                [-0.2921184613487701],
-                [0.20088114454313125],
-            ],
-        ],
-        "regressor.intercepts_": [
-            [
-                0.5652480230183061,
-                -0.14461022167946497,
-                0.0734880322671333,
-                0.774381965871222,
-                -0.5810353861392875,
-                -0.6447073648626959,
-                0.6469258828055443,
-                -0.8684424314182713,
-                0.42510685510971963,
-                -0.0666156988876174,
-            ],
-            [0.12043980294087006],
-        ],
-    },
-}
-
-
-def main_features(model, res):
-    return [
-        res.main_chain_correlation,
-        1 if res.has_pepflip_peak else 0,
-        max(atom.max_overlap for atom in res.main_chain_atoms),
-        max(atom.bfactor_zscore for atom in res.main_chain_atoms),
-        max(atom.bchange_zscore for atom in res.main_chain_atoms),
-        mean([atom.density_zscore for atom in res.main_chain_atoms]),
-        min(atom.diff_zscore for atom in res.main_chain_atoms),
-        min(atom.density_zscore for atom in res.main_chain_atoms),
-        None if res.next is None else res.next.atoms[" CA "].diff_zscore,
-        res.ramachandran_score,
-        model.resolution,
-        res.twistedness,
-    ]
-
-
-def side_features(model, res):
-    return [
-        res.side_chain_correlation,
-        max(atom.max_overlap for atom in res.side_chain_atoms),
-        max(atom.bfactor_zscore for atom in res.side_chain_atoms),
-        max(atom.bchange_zscore for atom in res.side_chain_atoms),
-        mean([atom.density_zscore for atom in res.side_chain_atoms]),
-        min(atom.diff_zscore for atom in res.side_chain_atoms),
-        min(atom.density_zscore for atom in res.side_chain_atoms),
-        model.resolution,
-        res.rotamer_score,
-    ]
-
-
-def add_medians(X, medians):
-    return [medians[i] if X[i] is None else X[i] for i in range(len(X))]
-
-
-def scale(X, means, scales):
-    return [(X[i] - means[i]) / scales[i] for i in range(len(X))]
-
-
-def predict(X, coefs, intercepts):
-    n_layers = len(coefs) + 1
-    values = [None] * n_layers
-    values[0] = list(X)
-    for i in range(1, n_layers):
-        values[i] = list(intercepts[i - 1])
-        for j in range(len(intercepts[i - 1])):
-            for k in range(len(values[i - 1])):
-                values[i][j] += values[i - 1][k] * coefs[i - 1][k][j]
-            if i < n_layers - 1:
-                values[i][j] = math.tanh(values[i][j])
-    return values[-1][0]
-
-
-# ATOM / RESIDUE / CHAIN / MODEL
-
-
-class Atom:
-    def __init__(self, model, residue, atom_info):
-        self.name = atom_info[0][0]
-        self.alt_conf = atom_info[0][1]
-        self.occupancy = atom_info[1][0]
-        self.bfactor = atom_info[1][1]
-        self.element = atom_info[1][2].strip()
-        self.atomic_number = atomic_numbers[self.element]
-        self.is_main_chain = self.name in main_chain_atoms
-        self.is_side_chain = not self.is_main_chain
-        self.point = atom_info[2]
-        self.density = density_at_point(model.imap, *self.point)
-        self.density_norm = self.density / self.atomic_number
-        self.diff_density = density_at_point(model.imap_diff, *self.point)
-        self.diff_norm = self.diff_density / self.atomic_number
-        self.max_overlap = 0
-
-
-class Residue:
-    def __init__(self, model, spec, name):
-        self.spec = spec
-        self.name = name
-        self.chain = spec[0]
-        self.resno = spec[1]
-        self.ins_code = spec[2]
-        self.next = None
-        self.prev = None
-        self.twistedness = None
-        self.ramachandran_score = None
-        self.rotamer_score = rotamer_score(model.imol, spec[0], spec[1], spec[2], "")
-        self.atoms = {}
-        for atom_info in residue_info_py(model.imol, *spec):
-            atom = Atom(model, self, atom_info)
-            model.atoms.append(atom)
-            self.atoms[atom.name] = atom
-        self.main_chain_atoms = [atom for atom in self.atoms.values() if atom.is_main_chain]
-        self.side_chain_atoms = [atom for atom in self.atoms.values() if atom.is_side_chain]
-        self.truncatable = len(self.side_chain_atoms) > 0
-
-
-class Chain:
-    def __init__(self):
-        self.residues = []
-        self.correctnesses = []
-
-
-class Model:
-    def __init__(self, imol, imap, imap_diff):
-        self.imol = imol
-        self.imap = imap
-        self.imap_diff = imap_diff
-        self.resolution = data_resolution(self.imap)
-        self.residues = []
-        self.residue_spec_dict = {}
-        self.atoms = []
-        for ichain in range(n_chains(imol)):
-            chain_id = chain_id_py(imol, ichain)
-            for serial_num in range(chain_n_residues(chain_id, imol)):
-                resno = seqnum_from_serial_number(imol, chain_id, serial_num)
-                ins_code = insertion_code_from_serial_number(imol, chain_id, serial_num)
-                spec = [chain_id, resno, ins_code]
-                name = residue_name(imol, *spec)
-                if name in protein_residues:
-                    residue = Residue(self, spec, name)
-                    self.residues.append(residue)
-                    self.residue_spec_dict[tuple(spec)] = residue
-        self.set_connections()
-        self.set_correlations()
-        self.set_bond_changes()
-        self.set_zscores()
-        self.set_ramachandran()
-        self.set_overlaps()
-        self.set_pepflip_peaks()
-        self.set_correctness()
-        self.set_chains()
-
-    def set_connections(self):
-        for i in range(len(self.residues) - 1):
-            res1 = self.residues[i]
-            res2 = self.residues[i + 1]
-            if " CA " not in res1.atoms or " C  " not in res1.atoms:
-                continue
-            if " N  " not in res2.atoms or " CA " not in res2.atoms:
-                continue
-            point1 = res1.atoms[" C  "].point
-            point2 = res2.atoms[" N  "].point
-            if distance(point1, point2) < 1.7:
-                res1.next = res2
-                res2.prev = res1
-                xyz1 = res1.atoms[" CA "].point
-                xyz2 = res1.atoms[" C  "].point
-                xyz3 = res2.atoms[" N  "].point
-                xyz4 = res2.atoms[" CA "].point
-                omega = abs(torsion(xyz1, xyz2, xyz3, xyz4))
-                twistedness = min(omega, 180 - omega)
-                if twistedness < 0:
-                    raise Exception("Negative twistedness between %s and %s" % (res1.spec, res2.spec))
-                res1.twistedness = max(twistedness, 0 if res1.twistedness is None else res1.twistedness)
-                res2.twistedness = max(twistedness, 0 if res2.twistedness is None else res2.twistedness)
-
-    def set_correlations(self):
-        self.set_correlation("main_chain_correlation", 1)
-        self.set_correlation("side_chain_correlation", 3)
-
-    def set_bond_changes(self):
-        for residue in self.residues:
-            for atom in residue.atoms.values():
-                atom.bchange = -99999
-                for attached in attached_atoms(atom, residue):
-                    bchange = (atom.bfactor - attached.bfactor) / attached.bfactor
-                    atom.bchange = max(bchange, atom.bchange)
-                if atom.bchange == -99999:
-                    raise Exception("Residue %s atom %s is not bonded to anything" % (residue.spec, atom.name))
-
-    def set_correlation(self, attr, mask):
-        specs = [residue.spec for residue in self.residues]
-        for correlation in map_to_model_correlation_per_residue_py(self.imol, specs, mask, self.imap):
-            residue = self.residue_spec_dict[tuple(correlation[0][1:])]
-            setattr(residue, attr, correlation[1])
-
-    # Uses a modified Z-score calculated using medians
-    # Boris Iglewicz and David Hoaglin
-    # Volume 16: How to Detect and Handle Outliers
-    # The ASQC Basic References in Quality Control: Statistical Techniques
-    # Edited by Edward F. Mykytka, Ph.D.
-    # Page 11
-    # 1993
-    def set_zscores(self):
-        def set_zscore(attr, zscore_attr):
-            main_values = [getattr(atom, attr) for atom in self.atoms if atom.is_main_chain]
-            side_values = [getattr(atom, attr) for atom in self.atoms if atom.is_side_chain]
-            main_median = median(main_values)
-            side_median = median(side_values)
-            main_mad = median_absolute_deviation(main_values)
-            side_mad = median_absolute_deviation(side_values)
-            for residue in self.residues:
-                for atom in residue.atoms.values():
-                    value = getattr(atom, attr)
-                    if atom.is_main_chain:
-                        zscore = 0 if main_mad == 0 else 0.6745 * (value - main_median) / main_mad
-                    else:
-                        zscore = 0 if side_mad == 0 else 0.6745 * (value - side_median) / side_mad
-                    setattr(atom, zscore_attr, zscore)
-
-        set_zscore("density_norm", "density_zscore")
-        set_zscore("diff_norm", "diff_zscore")
-        set_zscore("bfactor", "bfactor_zscore")
-        set_zscore("bchange", "bchange_zscore")
-
-    def set_ramachandran(self):
-        for item in all_molecule_ramachandran_score_py(self.imol)[5]:
-            spec = item[1][1:]
-            if tuple(spec) in self.residue_spec_dict:
-                residue = self.residue_spec_dict[tuple(spec)]
-                residue.ramachandran_score = item[2]
-
-    def set_overlaps(self):
-        for overlap in molecule_atom_overlaps_py(self.imol):
-            for n in (1, 2):
-                spec = tuple(overlap["atom-%d-spec" % n][1:4])
-                if spec in self.residue_spec_dict:
-                    atom_name = overlap["atom-%d-spec" % n][4]
-                    atom = self.residue_spec_dict[spec].atoms[atom_name]
-                    atom.max_overlap = max(atom.max_overlap, overlap["overlap-volume"])
-
-    def set_pepflip_peaks(self):
-        peaks = map_peaks_around_molecule_py(self.imap_diff, 4.62567528, False, self.imol)
-
-        def has_o_moving_peak(residue):
-            if residue.next is None:
-                return False
-            if " O  " not in residue.atoms:
-                return False
-            ca1 = residue.atoms[" CA "].point
-            c = residue.atoms[" C  "].point
-            o = residue.atoms[" O  "].point
-            ca2 = residue.next.atoms[" CA "].point
-            for peak in peaks:
-                point = peak[1]
-                cd = distance(point, c)
-                if cd < 1.09769614 or cd > 3.22217608:
-                    continue
-                if angle(o, c, point) < 75.45064459:
-                    continue
-                ca1d = distance(point, ca1)
-                if ca1d < 0.66596873 or ca1d > 2.8910501:
-                    continue
-                ca2d = distance(point, ca2)
-                if ca2d < 2.23908466 or ca2d > 3.84813939:
-                    continue
-                return True
-            return False
-
-        def has_n_moving_peak(residue):
-            if residue.prev is None or not residue.prev.has_pepflip_peak:
-                return False
-            if " O  " not in residue.prev.atoms:
-                return False
-            o = residue.prev.atoms[" O  "].point
-            ca = residue.atoms[" CA "].point
-            for peak in peaks:
-                point = peak[1]
-                if distance(point, o) > 0.52516368:
-                    continue
-                if distance(point, ca) > 2.76397302:
-                    continue
-                return True
-            return False
-
-        for residue in self.residues:
-            residue.has_pepflip_peak = has_o_moving_peak(residue) or has_n_moving_peak(residue)
-
-    def set_correctness(self):
-        if len(training_data) == 0:
-            return
-        for res in self.residues:
-            X_main = main_features(self, res)
-            X_main = add_medians(X_main, training_data["main"]["medians"])
-            X_main = scale(X_main, training_data["main"]["scaler.mean_"], training_data["main"]["scaler.scale_"],)
-            res.main_chain_correctness = predict(
-                X_main, training_data["main"]["regressor.coefs_"], training_data["main"]["regressor.intercepts_"],
-            )
-            if res.truncatable:
-                X_side = side_features(self, res)
-                X_side = add_medians(X_side, training_data["side"]["medians"])
-                X_side = scale(X_side, training_data["side"]["scaler.mean_"], training_data["side"]["scaler.scale_"],)
-                res.side_chain_correctness = predict(
-                    X_side, training_data["side"]["regressor.coefs_"], training_data["side"]["regressor.intercepts_"],
-                )
-
-    def set_chains(self):
-        self.chains = {}
-        for residue in self.residues:
-            chain_id = residue.chain
-            if chain_id not in self.chains:
-                self.chains[chain_id] = Chain()
-            chain = self.chains[chain_id]
-            chain.residues.append(residue)
-            if hasattr(residue, "main_chain_correctness"):
-                chain.correctnesses.append(residue.main_chain_correctness)
-        for chain in self.chains.values():
-            if len(chain.correctnesses) > 0:
-                chain.correctness = mean(chain.correctnesses)
-
-
-# SCRIPTING
-
-
-def prune(
-    imol,
-    imap,
-    imap_diff,
-    chains=True,
-    chain_threshold="auto",
-    max_chain_fraction=0.2,
-    max_chain_length=20,
-    residues=True,
-    residue_threshold="auto",
-    max_residue_fraction=0.2,
-    remove_isolated_residues=True,
-    sidechains=True,
-    sidechain_threshold="auto",
-    max_sidechain_fraction=0.2,
-):
-
-    model = Model(imol, imap, imap_diff)
-
-    if len(model.chains) < 1:
-        return
-
-    if chains:
-        main_median = median([r.main_chain_correctness for r in model.residues])
-        if chain_threshold == "auto":
-            chain_threshold = main_median * 0.2
-        print(
-            "ML_CORRECTNESS: Deleting chains (up to %d residues long) with scores < %.3f"
-            % (max_chain_length, chain_threshold)
-        )
-        print("ML_CORRECTNESS: Up to %.0f%% of residues will be deleted" % (max_chain_fraction * 100))
-        max_deleted = len(model.residues) * max_chain_fraction
-        deleted = 0
-        remaining = []
-        for chain_id in sorted(model.chains,):
-            chain = model.chains[chain_id]
-            if (
-                chain.correctness < chain_threshold
-                and len(chain.residues) <= max_chain_length
-                and deleted + len(chain.residues) <= max_deleted
-            ):
-                deleted += len(chain.residues)
-                delete_chain(imol, chain_id)
-            else:
-                remaining.extend([r for r in chain.residues])
-        print("ML_CORRECTNESS: Deleted %.0f%% of residues" % (float(deleted) / len(model.residues) * 100))
-    else:
-        remaining = [r for r in model.residues]
-
-    if len(remaining) < 1:
-        return
-
-    if residues:
-        main_median = median([r.main_chain_correctness for r in remaining])
-        if residue_threshold == "auto":
-            residue_threshold = main_median * 0.5
-        print("ML_CORRECTNESS: Deleting residues with scores < %.3f" % residue_threshold)
-        print("ML_CORRECTNESS: Up to %.0f%% of residues will be deleted" % (max_residue_fraction * 100))
-        max_deleted = len(remaining) * max_residue_fraction
-        deleted = 0
-        remaining.sort(key=lambda r: r.main_chain_correctness)
-        for residue in remaining:
-            residue.delete = False
-            if residue.main_chain_correctness < residue_threshold and deleted + 1 <= max_deleted:
-                deleted += 1
-                residue.delete = True
-        if remove_isolated_residues:
-            for residue in remaining:
-                if residue.prev is not None and not residue.prev.delete:
-                    continue
-                if residue.next is not None and not residue.next.delete:
-                    continue
-                residue.delete = True
-        for residue in remaining:
-            if residue.delete:
-                delete_residue(imol, residue.chain, residue.resno, residue.ins_code)
-        print("ML_CORRECTNESS: Deleted %.0f%% of residues" % (float(deleted) / len(remaining) * 100))
-        remaining = [r for r in remaining if not r.delete]
-
-    remaining = [r for r in remaining if r.truncatable]
-    if len(remaining) < 1:
-        return
-
-    if sidechains:
-        side_median = median([r.side_chain_correctness for r in model.residues if r.truncatable])
-        if sidechain_threshold == "auto":
-            sidechain_threshold = side_median * 0.5
-        print("ML_CORRECTNESS: Deleting sidechains with scores < %.3f" % sidechain_threshold)
-        print("ML_CORRECTNESS: Up to %.0f%% of sidechains will be deleted" % (max_sidechain_fraction * 100))
-        max_deleted = len(remaining) * max_sidechain_fraction
-        deleted = 0
-        remaining.sort(key=lambda r: r.side_chain_correctness)
-        for residue in remaining:
-            if residue.side_chain_correctness < sidechain_threshold and deleted + 1 < max_deleted:
-                deleted += 1
-                delete_residue_sidechain(imol, residue.chain, residue.resno, residue.ins_code, 0)
-        print("ML_CORRECTNESS: Deleted %.0f%% of sidechains" % (float(deleted) / len(remaining) * 100))