mhcseqs 2.2.2__tar.gz → 2.3.0__tar.gz

{mhcseqs-2.2.2 → mhcseqs-2.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mhcseqs
-Version: 2.2.2
+Version: 2.3.0
 Summary: MHC sequence curation and binding groove extraction
 License-Expression: Apache-2.0
 Classifier: Development Status :: 4 - Beta
@@ -107,7 +107,7 @@ Three sources are merged into a single dataset:
 | UniProt | 20,566 | 500+ species | Curated diverse MHC, B2M, H-2 references (shipped in package) |
 | **Total raw** | **77,576** | | |
 | **After merge/dedup** | **55,696** | | One representative per two-field allele |
-| **Groove OK** | **54,164** | | 97.2% of representatives |
+| **Groove OK** | **54,155** | | 97.2% of representatives |
 
 ### By species category
 
@@ -115,13 +115,13 @@ Three sources are merged into a single dataset:
 |---|---:|---:|---:|---:|
 | human | 25,364 | 99.8% | 17,364 | 5,347 |
 | nhp | 7,125 | 98.0% | 4,582 | 1,429 |
-| bird | 9,312 | 99.0% | 549 | 562 |
+| bird | 9,312 | 99.1% | 553 | 558 |
 | fish | 4,859 | 97.2% | 777 | 1,259 |
 | ungulate | 1,768 | 96.2% | 603 | 303 |
-| murine | 1,625 | 89.7% | 509 | 123 |
+| murine | 1,625 | 89.0% | 466 | 158 |
 | carnivore | 484 | 99.6% | 164 | 0 |
-| other_mammal | 943 | 95.1% | 281 | 146 |
-| other_vertebrate | 1,137 | 96.1% | 325 | 131 |
+| other_mammal | 943 | 95.0% | 275 | 152 |
+| other_vertebrate | 1,137 | 96.2% | 327 | 129 |
 
 "Groove OK" includes both full-length and fragment parses. "Full" means both groove
 halves present (class I: α1 + α2, ~183 aa; class II: single chain's groove half with

{mhcseqs-2.2.2 → mhcseqs-2.3.0}/README.md

@@ -86,7 +86,7 @@ Three sources are merged into a single dataset:
 | UniProt | 20,566 | 500+ species | Curated diverse MHC, B2M, H-2 references (shipped in package) |
 | **Total raw** | **77,576** | | |
 | **After merge/dedup** | **55,696** | | One representative per two-field allele |
-| **Groove OK** | **54,164** | | 97.2% of representatives |
+| **Groove OK** | **54,155** | | 97.2% of representatives |
 
 ### By species category
 
@@ -94,13 +94,13 @@ Three sources are merged into a single dataset:
 |---|---:|---:|---:|---:|
 | human | 25,364 | 99.8% | 17,364 | 5,347 |
 | nhp | 7,125 | 98.0% | 4,582 | 1,429 |
-| bird | 9,312 | 99.0% | 549 | 562 |
+| bird | 9,312 | 99.1% | 553 | 558 |
 | fish | 4,859 | 97.2% | 777 | 1,259 |
 | ungulate | 1,768 | 96.2% | 603 | 303 |
-| murine | 1,625 | 89.7% | 509 | 123 |
+| murine | 1,625 | 89.0% | 466 | 158 |
 | carnivore | 484 | 99.6% | 164 | 0 |
-| other_mammal | 943 | 95.1% | 281 | 146 |
-| other_vertebrate | 1,137 | 96.1% | 325 | 131 |
+| other_mammal | 943 | 95.0% | 275 | 152 |
+| other_vertebrate | 1,137 | 96.2% | 327 | 129 |
 
 "Groove OK" includes both full-length and fragment parses. "Full" means both groove
 halves present (class I: α1 + α2, ~183 aa; class II: single chain's groove half with

{mhcseqs-2.2.2 → mhcseqs-2.3.0}/mhcseqs/pipeline.py

@@ -20,6 +20,7 @@ from .alleles import (
     normalize_allele_name,
     normalize_mhc_class,
     parse_allele_name,
+    parse_gene_class,
 )
 from .domain_grammar import NON_MHC_ACCESSIONS
 from .domain_parsing import (
@@ -427,6 +428,16 @@ def _load_diverse_mhc_references() -> List[dict]:
             if not species_category:
                 species_category = normalize_mhc_species(organism) or ""
 
+            # Override mhc_class from mhcgnomes when the CSV value is
+            # wrong or unknown — UniProt annotations frequently misclassify
+            # RT1/H-2 class II genes (DOb, Ba, Eb, DQA, etc.) as class I.
+            if gene:
+                pgc = parse_gene_class(gene)
+                if pgc and pgc.get("mhc_class") in ("I", "II"):
+                    mhc_class = pgc["mhc_class"]
+                    if pgc.get("chain"):
+                        chain = pgc["chain"]
+
             # Fix chain for class II unknowns: try class_ii_alpha_gene check
             if mhc_class == "II" and chain in ("unknown", ""):
                 if gene and is_class_ii_alpha_gene(gene):

mhcseqs-2.3.0/mhcseqs/version.py

@@ -0,0 +1 @@
+__version__ = "2.3.0"

{mhcseqs-2.2.2 → mhcseqs-2.3.0}/mhcseqs.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mhcseqs
-Version: 2.2.2
+Version: 2.3.0
 Summary: MHC sequence curation and binding groove extraction
 License-Expression: Apache-2.0
 Classifier: Development Status :: 4 - Beta
@@ -107,7 +107,7 @@ Three sources are merged into a single dataset:
 | UniProt | 20,566 | 500+ species | Curated diverse MHC, B2M, H-2 references (shipped in package) |
 | **Total raw** | **77,576** | | |
 | **After merge/dedup** | **55,696** | | One representative per two-field allele |
-| **Groove OK** | **54,164** | | 97.2% of representatives |
+| **Groove OK** | **54,155** | | 97.2% of representatives |
 
 ### By species category
 
@@ -115,13 +115,13 @@ Three sources are merged into a single dataset:
 |---|---:|---:|---:|---:|
 | human | 25,364 | 99.8% | 17,364 | 5,347 |
 | nhp | 7,125 | 98.0% | 4,582 | 1,429 |
-| bird | 9,312 | 99.0% | 549 | 562 |
+| bird | 9,312 | 99.1% | 553 | 558 |
 | fish | 4,859 | 97.2% | 777 | 1,259 |
 | ungulate | 1,768 | 96.2% | 603 | 303 |
-| murine | 1,625 | 89.7% | 509 | 123 |
+| murine | 1,625 | 89.0% | 466 | 158 |
 | carnivore | 484 | 99.6% | 164 | 0 |
-| other_mammal | 943 | 95.1% | 281 | 146 |
-| other_vertebrate | 1,137 | 96.1% | 325 | 131 |
+| other_mammal | 943 | 95.0% | 275 | 152 |
+| other_vertebrate | 1,137 | 96.2% | 327 | 129 |
 
 "Groove OK" includes both full-length and fragment parses. "Full" means both groove
 halves present (class I: α1 + α2, ~183 aa; class II: single chain's groove half with

mhcseqs-2.2.2/mhcseqs/version.py

@@ -1 +0,0 @@
-__version__ = "2.2.2"