From 68ad8fb62a76a63c9bb802b4b8215c837df9adfe Mon Sep 17 00:00:00 2001
From: Stefan Bienert <stefan.bienert@unibas.ch>
Date: Mon, 3 Oct 2022 14:10:59 +0200
Subject: [PATCH] Add selection step
---
translate2modelcif.py | 116 +++++++++++++++++++-----------------------
1 file changed, 53 insertions(+), 63 deletions(-)
diff --git a/translate2modelcif.py b/translate2modelcif.py
index db57ff1..aa90588 100644
--- a/translate2modelcif.py
+++ b/translate2modelcif.py
@@ -249,7 +249,6 @@ def _check_model_extra_files_present(model_dir, pdb_file):
def _get_audit_authors():
"""Return the list of authors that produced this model."""
- # ToDo: tell Xabi that his name can't have a รก in mmCIF
return (
"Bartolec, T.",
"Vazquez-Campos, X.",
@@ -289,7 +288,8 @@ def _parse_colabfold_config(cnfg_file):
use_msa = False
elif cf_config["msa_mode"] == "custom":
print(
- "WARNING: Custom MSA mode used. Not clear from config what to do here!"
+ "WARNING: Custom MSA mode used. Not clear from config what to do "
+ + "here!"
)
seq_dbs = []
use_mmseqs = False
@@ -311,63 +311,71 @@ def _parse_colabfold_config(cnfg_file):
else:
raise ValueError(f"Unknown model_type {cf_config['model_type']}")
- # write description
- description = f"Model generated using ColabFold v{cf_config['version']}"
+ # write modeling description
+ mdl_description = f"Model generated using ColabFold v{cf_config['version']}"
if use_multimer:
- description += f" with AlphaFold-Multimer (v{multimer_version})"
+ mdl_description += f" with AlphaFold-Multimer (v{multimer_version})"
else:
- description += " with AlphaFold"
+ mdl_description += " with AlphaFold"
if cf_config["stop_at_score"] < 100:
# early stopping feature of ColabFold
upto = "up to "
else:
upto = ""
- description += (
+ mdl_description += (
f" producing {upto}{cf_config['num_models']} models"
f" with {upto}{cf_config['num_recycles']} recycles each"
)
if cf_config["use_amber"]:
- description += ", with AMBER relaxation"
+ mdl_description += ", with AMBER relaxation"
else:
- description += ", without model relaxation"
+ mdl_description += ", without model relaxation"
if cf_config["use_templates"]:
print(
"WARNING: ColabFold may use PDB70 or custom templates. "
"Not clear from config!"
)
- description += ", using templates"
+ mdl_description += ", using templates"
else:
- description += ", without templates"
+ mdl_description += ", without templates"
if cf_config["rank_by"] == "plddt":
- description += ", ranked by pLDDT"
+ mdl_description += ", ranked by pLDDT"
elif cf_config["rank_by"] == "ptmscore":
- description += ", ranked by pTM"
+ mdl_description += ", ranked by pTM"
elif cf_config["rank_by"] == "multimer":
- description += ", ranked by ipTM*0.8+pTM*0.2"
+ mdl_description += ", ranked by ipTM*0.8+pTM*0.2"
else:
raise ValueError(f"Unknown rank_by {cf_config['rank_by']}")
if use_msa:
- description += ", starting from"
+ mdl_description += ", starting from"
if use_mmseqs:
msa_type = "MSA"
else:
msa_type = "custom MSA"
if use_multimer:
if cf_config["pair_mode"] == "unpaired+paired":
- description += f" paired and unpaired {msa_type}s"
+ mdl_description += f" paired and unpaired {msa_type}s"
elif cf_config["pair_mode"] == "paired":
- description += f" paired {msa_type}s"
+ mdl_description += f" paired {msa_type}s"
elif cf_config["pair_mode"] == "unpaired":
- description += f" unpaired {msa_type}s"
+ mdl_description += f" unpaired {msa_type}s"
else:
raise ValueError(f"Unknown pair_mode {cf_config['pair_mode']}")
else:
- description += f" an {msa_type}"
+ mdl_description += f" an {msa_type}"
if use_mmseqs:
- description += f" from MMseqs2 ({'+'.join(seq_dbs)})"
+ mdl_description += f" from MMseqs2 ({'+'.join(seq_dbs)})"
else:
- description += " without an MSA"
- description += "."
+ mdl_description += " without an MSA"
+ mdl_description += "."
+
+ # write selection description
+ slct_description = (
+ "Select best model, which is either the top-ranked model as "
+ + "determined by the ColabFold pipeline (ipTM*0.8+pTM*0.2), or else "
+ + "the model with best congruence with crosslinks reported in the "
+ + "related study."
+ )
return {
"config": cf_config,
@@ -376,7 +384,8 @@ def _parse_colabfold_config(cnfg_file):
"use_msa": use_msa,
"use_multimer": use_multimer,
"multimer_version": multimer_version,
- "description": description,
+ "modeling_description": mdl_description,
+ "selection_description": slct_description,
}
@@ -388,7 +397,7 @@ def _get_protocol_steps_and_software(config_data):
step = {
"method_type": "modeling",
"name": None,
- "details": config_data["description"],
+ "details": config_data["modeling_description"],
}
# get input data
# Must refer to data already in the JSON, so we try keywords
@@ -415,21 +424,7 @@ def _get_protocol_steps_and_software(config_data):
"classification": "data collection",
"description": "Many-against-Many sequence searching",
# ToDo: add citation to ihm.citations
- "citation": ihm.Citation(
- pmid="30615063",
- title="MMseqs2 desktop and local web server app for fast, "
- + "interactive sequence searches.",
- journal="Bioinformatics",
- volume=35,
- page_range=(2856, 2858),
- year=2019,
- authors=[
- "Mirdita, M.",
- "Steinegger, M.",
- "Soeding, J.",
- ],
- doi="10.1093/bioinformatics/bty1057",
- ),
+ "citation": ihm.citations.mmseqs2,
"location": "https://github.com/soedinglab/mmseqs2",
"type": "package",
"version": None,
@@ -496,29 +491,22 @@ def _get_protocol_steps_and_software(config_data):
protocol.append(step)
# model selection step
- # ToDo [input/ internal]: model selection step on a single model is a bit
- # silly, how do we get a list of models?
- # GT-NOTES:
- # - input/output should be ok without list of models
- # - rank of model is already stored in _ma_model_list.model_name and
- # _ma_data.name (in _store_as_modelcif)
- # - ColabFold ranking details is already in details of step above.
- # - Suggestion: add extra step only if AF-ranking was overruled and
- # include it in step above.
-
- # step = {
- # "method_type": "model selection",
- # "name": "ma_protocol_step.step_name",
- # "details": "Select best model, which is either the top-ranked model "
- # + "as determined by the ColabFold pipeline "
- # + "(ipTM*0.8+pTM*0.2), or else the model with best "
- # + "congruence with crosslinks reported in the related study.",
- # }
- # step["input"] = "model"
- # step["output"] = "model"
- # step["software"] = []
- # step["software_parameters"] = {}
- # protocol.append(step)
+ if (
+ "selection_description" not in config_data
+ or len(config_data["selection_description"]) == 0
+ ):
+ return protocol
+
+ step = {
+ "method_type": "model selection",
+ "name": None,
+ "details": config_data["selection_description"],
+ }
+ step["input"] = "model"
+ step["output"] = "model"
+ step["software"] = []
+ step["software_parameters"] = {}
+ protocol.append(step)
return protocol
@@ -584,7 +572,9 @@ def _fetch_upkb_entry(up_ac):
data["up_organism"] = ""
data["up_sequence"] = ""
data["up_ac"] = up_ac
- rspns = requests.get(f"https://www.uniprot.org/uniprot/{up_ac}.txt")
+ rspns = requests.get(
+ f"https://www.uniprot.org/uniprot/{up_ac}.txt", timeout=180
+ )
for line in rspns.iter_lines(decode_unicode=True):
if line.startswith("ID "):
sline = line.split()
--
GitLab