From baa87ea048e6ec48caaadd026df2f937a56a2430 Mon Sep 17 00:00:00 2001
From: Matiq <kokagishi@proton.me>
Date: Tue, 8 Sep 2026 23:31:01 +0300
Subject: [PATCH] fix OncoKB endpoint byHgvsVariant->byGenomicChange (GRCh38)

- 7:g.140753336A>T now Tier I/Oncogenic (was 404)
- 1:g.18396307G>A correctly Unknown/empty (expected)
- offline still skips OncoKB as requested (VAF>20% default)
---
 annotate_clean.py | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

diff --git a/annotate_clean.py b/annotate_clean.py
index ac808f2..1d4722c 100644
--- a/annotate_clean.py
+++ b/annotate_clean.py
@@ -45,7 +45,7 @@ ONCO_COLORS = {
     "Resistance": "FFC7CE",
 }
 
-ONCOKB_URL = "https://www.oncokb.org/api/v1/annotate/mutations/byHgvsVariant"
+ONCOKB_URL = "https://www.oncokb.org/api/v1/annotate/mutations/byGenomicChange"
 CIVIC_URL = "https://civicdb.org/api/variants?count=10000"
 
 def load_token(path="~/.config/oncokb/token"):
@@ -202,7 +202,7 @@ def vep_annotate(df, reference, gtf_path=None, offline=False, all_transcripts=Fa
     return pd.DataFrame(rows)
 
 def fetch_oncokb(hgvs_g_list, token, tumor_type="All Solid Tumors", offline=False):
-    """Batch query OncoKB byHgvsVariant. Returns dict hgvs_g -> {amp, oncogenic}."""
+    """Batch query OncoKB byGenomicChange (GRCh38). Returns dict hgvs_g -> {amp, oncogenic}."""
     if not token or offline:
         return {}
     if len(hgvs_g_list) > 500:
@@ -214,7 +214,20 @@ def fetch_oncokb(hgvs_g_list, token, tumor_type="All Solid Tumors", offline=Fals
         if i % 100 == 0 and len(hgvs_g_list) > 100:
             print(f"  [oncokb] {i}/{len(hgvs_g_list)} ...", flush=True)
         try:
-            params = {"hgvsg": hgvs, "tumorType": tumor_type}
+            # hgvs like "7:g.140753336A>T" -> genomicLocation "7,140753336,140753336,A,T"
+            try:
+                chrom_part, rest = hgvs.split(":g.")
+                pos_ref, alt = rest.split(">")
+                # pos_ref like "140753336A"
+                pos_str = "".join(c for c in pos_ref if c.isdigit())
+                ref = "".join(c for c in pos_ref if c.isalpha())
+                pos = int(pos_str) if pos_str else 0
+            except Exception:
+                chrom_part, pos, ref, alt = "1", 0, "A", "T"
+                pos = 0
+            chrom = chrom_part
+            params = {"genomicLocation": f"{chrom},{pos},{pos},{ref},{alt}",
+                      "referenceGenome": "GRCh38", "tumorType": tumor_type}
             resp = requests.get(ONCOKB_URL, headers=headers, params=params, timeout=(3, 5))
             if resp.status_code == 200:
                 j = resp.json()