feat(report): flag a gene hit twice across engines, e.g. a deletion plus a variant on the other copy

assets/acmg_sf.tsv

@@ -9,7 +9,10 @@
 #   report_if       when a P/LP variant (ClinVar, or a rare HIGH-impact record) is a secondary finding:
 #                     any             one variant is enough
 #                     biallelic       two variants: a homozygote, or two different variants even when
-#                                     their phase is unknown (Table 1 footnote a)
+#                                     their phase is unknown (Table 1 footnote a); a called deletion
+#                                     of the gene (AnnotSV, step 35's copy number) counts as one of the
+#                                     two unless read-backed phase puts it on the variant's copy
+#                                     (bin/collect_summary.py two_hits); so it does for hemizygous
 #                     hom_variant     only a variant of the variants column, homozygous
 #                     hemizygous      a hemizygous or homozygous variant, or two heterozygous ones; a
 #                                     single heterozygous variant is not reported

assets/carrier_genes.tsv

@@ -14,7 +14,9 @@
 # severity: alleles of the gene that are not a full loss of function, separated by ";": each is
 #   hypomorph or pseudodeficiency, then chrom:pos:ref:alt (GRCh38), then a name; "." for none. Such
 #   an allele never makes a carrier finding on its own (the couple screen, step 39, and the
-#   carrier list of the sample report).
+#   carrier list of the sample report). With a called deletion of the gene it is a second hit (the
+#   two-hit lines of the sample report, bin/collect_summary.py two_hits). In the couple screen a
+#   hypomorph beside a full loss of function in the other partner's same gene is a shared gene.
 # Rebuilt with: couple_report.py genes (docs/39-couple.md says how).
 gene	chrom	start	end	inheritance	lists	condition	severity
 ISG15	chr1	1001138	1014540	AR	Orphanet	Mendelian susceptibility to mycobacterial diseases due to complete ISG15 deficiency	.
@@ -186,7 +188,7 @@ HSD3B2	chr1	119413725	119453802	AR	Orphanet	Congenital adrenal hyperplasia due t
 PHGDH	chr1	119648411	119744227	AR	Orphanet	3-phosphoglycerate dehydrogenase deficiency, infantile/juvenile form; and 1 other	.
 HMGCS2	chr1	119747979	119769092	AR	Orphanet	3-hydroxy-3-methylglutaryl-CoA synthase deficiency	.
 PEX11B	chr1	145911348	145918747	AR	Orphanet	Infantile Refsum disease; and 2 others	.
-RBM8A	chr1	145920935	145928067	AR	Orphanet	Thrombocytopenia-absent radius syndrome	.
+RBM8A	chr1	145920935	145928067	AR	Orphanet	Thrombocytopenia-absent radius syndrome	hypomorph chr1:145927447:C:T c.-21G>A (rs139428292), a low-expression allele: TAR syndrome with a null allele such as the 1q21.1 deletion on the other copy; hypomorph chr1:145927328:C:G c.67+32G>C (rs201779890), the other low-expression allele of TAR syndrome
 HJV	chr1	146017468	146036746	AR	Orphanet	HJV or HAMP-related hemochromatosis	.
 GJA8	chr1	147902795	147909269	AR	Orphanet	Cataract-microcornea syndrome; and 3 others	.
 NOTCH2NLC	chr1	149390534	149471833	AR	Orphanet	Oculopharyngodistal myopathy	.

bin/couple_report.py

@@ -234,9 +234,12 @@ def common_match(benign, chrom, s0, e0):
 
 # --- findings ----------------------------------------------------------------
 
-def finding(gene, text, copies):
-    """copies: how many of the person's copies the finding is on (1 or 2)."""
-    return {"gene": gene, "text": text, "copies": copies}
+def finding(gene, text, copies, hypomorph=None):
+    """copies: how many of the person's copies the finding is on (1 or 2).
+    hypomorph: the name of a low-function allele of the list's severity
+    column, which counts only beside a full loss of function in the partner
+    (screen_couple)."""
+    return {"gene": gene, "text": text, "copies": copies, "hypomorph": hypomorph}
 
 
 def zygosity(gt, chrom, sex, par):
@@ -259,7 +262,11 @@ def stars_text(n):
 def clinvar_findings(path, genes, sex, par, aside=None):
     """Step 06's hits in genes of the list, and the count of all hits. A hit
     the list marks as a hypomorph or pseudodeficiency allele (its severity
-    column) is no carrier finding: it goes to aside as a text."""
+    column) is no carrier finding on its own: it goes to aside as a text. A
+    hypomorph is also returned as a finding marked hypomorph: beside a full
+    loss of function in the other partner's same gene it is the second hit of
+    the child (TAR syndrome is a null allele plus an RBM8A low-expression
+    allele); screen_couple drops it otherwise. total counts every hit."""
     out, total = [], 0
     for r in collect_summary.vcf_records(path):
         if len(r) < 8:
@@ -276,6 +283,10 @@ def clinvar_findings(path, genes, sex, par, aside=None):
                 if aside is not None:
                     aside.append("%s:%d %s>%s %s in %s: %s, a %s allele" % (
                         h["chrom"], h["pos"], h["ref"], h["alt"], z, g, sev["name"], sev["kind"]))
+                if sev["kind"] == "hypomorph":
+                    text = "%s:%d %s>%s %s, %s, a hypomorph allele" % (h["chrom"], h["pos"], h["ref"], h["alt"],
+                                                                       z, sev["name"])
+                    out.append(finding(g, text, 1 if z == "het" else 2, sev["name"]))
                 continue
             if g in genes:
                 ident = "" if r[2] in (".", "") else ", ClinVar %s" % r[2]
@@ -485,6 +496,11 @@ def ar_row(g, a, b, fa, fb):
     more = [n for n, f in ((a, fa), (b, fb)) if len(f) > 1 and all(x["copies"] == 1 for x in f)]
     why = ("Both partners carry a change in %s, linked to %s (autosomal recessive). A child who inherits "
            "the change of each parent can have the condition: %s." % (g["gene"], g["condition"], risk))
+    low = [n for n, f in ((a, fa), (b, fb)) if f and all(x.get("hypomorph") for x in f)]
+    if low:
+        why += (" The change of %s is a low-function (hypomorph) allele, not a full loss of function: with the "
+                "other parent's loss-of-function change a child can have the condition, for some genes in a "
+                "milder form; two such alleles alone are set aside." % " and ".join(low))
     if more:
         why += (" %s %s two changes in the gene; when they sit on different copies, every child gets one, "
                 "and the chance is higher." % (" and ".join(more), "has" if len(more) == 1 else "have"))
@@ -520,10 +536,13 @@ def xl_lines(g, who, sex, f):
 def screen_couple(a, b, sexes, found, genes):
     """Rows of the couple table, shared genes first, then X-linked, then the rest."""
     rows = []
-    for name in sorted(set(x["gene"] for x in found[a] + found[b])):
+    # A hypomorph (finding marked hypomorph) counts only when the other
+    # partner has a full loss of function in the same gene.
+    full = {s: set(x["gene"] for x in found[s] if not x.get("hypomorph")) for s in (a, b)}
+    for name in sorted(full[a] | full[b]):
         g = genes[name]
-        fa = [x for x in found[a] if x["gene"] == name]
-        fb = [x for x in found[b] if x["gene"] == name]
+        fa = [x for x in found[a] if x["gene"] == name and (not x.get("hypomorph") or name in full[b])]
+        fb = [x for x in found[b] if x["gene"] == name and (not x.get("hypomorph") or name in full[a])]
         if g["inheritance"] == "XL":
             parts = [xl_lines(g, w, sexes.get(w), f) for w, f in ((a, fa), (b, fb)) if f]
             result, risk, why = "x_linked", "; ".join(p[0] for p in parts), " ".join(p[1] for p in parts)
@@ -686,7 +705,7 @@ def cmd_screen(a):
                 gaps.append("Carrier deletion alleles of %s: no interval table (step 35)." % s)
             aside = []
             f, total = clinvar_findings(clinvar[s], genes, sexes[s], par, aside)
-            info = {"clinvar": "%d / %d" % (len(f), total)}
+            info = {"clinvar": "%d / %d" % (sum(1 for x in f if not x["hypomorph"]), total)}
             for note in aside:
                 gaps.append("Set aside for %s, not a carrier finding on its own: %s." % (s, note))
             if s not in delly:
@@ -860,7 +879,9 @@ def place(g):
         "# severity: alleles of the gene that are not a full loss of function, separated by \";\": each is",
         "#   hypomorph or pseudodeficiency, then chrom:pos:ref:alt (GRCh38), then a name; \".\" for none. Such",
         "#   an allele never makes a carrier finding on its own (the couple screen, step 39, and the",
-        "#   carrier list of the sample report).",
+        "#   carrier list of the sample report). With a called deletion of the gene it is a second hit (the",
+        "#   two-hit lines of the sample report, bin/collect_summary.py two_hits). In the couple screen a",
+        "#   hypomorph beside a full loss of function in the other partner's same gene is a shared gene.",
         "# Rebuilt with: couple_report.py genes (docs/39-couple.md says how).",
     ]
     with open(a.out, "w", encoding="utf-8") as f:

bin/render_report.py

@@ -1635,7 +1635,7 @@ def sf_rows(sf):
     the same in both reports: ClinVar hits, rare HIGH-impact records, then
     class 4 and 5 SVs."""
     def extra(x):
-        return "".join(f"; {v}" for v in (x.get("severity"), x.get("rule_note")) if v)
+        return "".join(f"; {v}" for v in (x.get("severity"), x.get("rule_note"), x.get("two_hit")) if v)
     rows = [(x["gene"], x["variant"], x["genotype"],
              "ClinVar " + x["significance"] + ("; low-confidence call: check the reads" if x.get("low_confidence") else "")
              + extra(x)) for x in sf["clinvar_hits"]]
@@ -1655,8 +1655,9 @@ def sf_apart_lines(sf):
     no rule could be applied."""
     if sf.get("unavailable"):
         return [f"Secondary findings: not assessed, {sf['unavailable']}"]
-    return [f"{x['gene']} {x['variant']} {x.get('genotype') or '.'} ({x['source']}): not reported, {x['reason']}"
-            for x in sf.get("not_reported", [])]
+    return ([f"{x['gene']} {x['variant']} {x.get('genotype') or '.'} ({x['source']}): not reported, {x['reason']}"
+             for x in sf.get("not_reported", [])]
+            + ([sf["two_hit_note"]] if sf.get("two_hit_note") else []))
 
 
 def carrier_lines(c):
@@ -1665,12 +1666,31 @@ def carrier_lines(c):
         return [f"Carrier list: not made, {c['unavailable']}"]
     lines = [f"Carrier list ({c['genes_on_list']} genes): {len(c['carriers'])} finding(s)"
              + (f", {len(c['set_aside'])} set aside" if c["set_aside"] else "")]
+    def hemi(x):
+        return f" ({x['hemizygous']})" if x.get("hemizygous") else ""
     for x in c["carriers"]:
-        lines.append(f"{x['gene']} ({x['condition']}; {x['inheritance']}): {x['variant']} {x['genotype']}, "
+        lines.append(f"{x['gene']} ({x['condition']}; {x['inheritance']}): {x['variant']} {x['genotype']}{hemi(x)}, "
                      f"ClinVar {x['significance']}")
     for x in c["set_aside"]:
-        lines.append(f"Set aside: {x['gene']} {x['variant']} {x['genotype']}: {x['name']} is a {x['kind']} allele, "
-                     "not a carrier finding on its own")
+        lines.append(f"Set aside: {x['gene']} {x['variant']} {x['genotype']}{hemi(x)}: {x['name']} is a {x['kind']} "
+                     "allele, not a carrier finding on its own")
+    return lines + two_hit_lines(c.get("two_hits"))
+
+
+def two_hit_lines(t):
+    """The genes a called deletion and a ClinVar hit both reach (collect_summary.two_hits), and which deletion
+    callers were read; none for a summary written before the rule."""
+    if not t:
+        return []
+    lines = []
+    for x in t["genes"]:
+        allele = f" ({x['allele']})" if x.get("allele") else ""
+        lines.append(f"{x['verdict'].capitalize()}: {x['gene']} ({x['condition']}; AR), both copies may be hit: "
+                     f"{x['deletion']} and {x['variant']} {x['genotype']}{allele}; {x['why']}")
+    lines.append("Two hits across callers (a deletion plus a variant on the other copy): "
+                 + (f"{len(t['genes'])} gene(s); " if t["genes"] else "none found; ")
+                 + "deletions read from " + ("; ".join(t["deletions_read"]) or "no caller")
+                 + (f"; not read: {'; '.join(t['deletions_not_read'])}" if t["deletions_not_read"] else ""))
     return lines
 
 

docs/interpreting-results.md

@@ -76,7 +76,15 @@ The most common "pathogenic" finding in any genome is **heterozygous carrier sta
 - BTD with p.Asp444His: reported as partial, not profound, biotinidase deficiency;
 - AnnotSV class 4 and 5 structural variants (step 5) in BRCA1, BRCA2, MLH1, MSH2 and LDLR reach the same card, and so does a deletion of the 3' end of EPCAM (exons 8 and 9), which silences MSH2 (Kovacs et al., Hum Mutat 2009;30(2):197-203).
 
-A deletion in trans with a small variant in a recessive gene is not counted as two variants yet. The TTN rule (truncating variants only) and the LMNA rule (missense only with cardiac evidence) are printed beside the finding, not applied. It is still a list to review with a clinician.
+**Two hits across callers: a deletion plus a variant on the other copy.** Small-variant callers see one hit when the other copy of the gene is deleted. So the carrier list also joins the deletions the other engines call with the ClinVar hits, gene by gene (`bin/collect_summary.py`, `two_hits`). The deletions come from AnnotSV's class 4 and 5 deletions (step 5), step 35's interval table (the recurrent regions of `bin/cn_loci.tsv`, such as 1q21.1), its HBA1/HBA2 count and Parascopy's SMN1 count. A recessive gene that a deletion and a hit both reach gets one line:
+
+- **biallelic** when the reads show the two hits on different copies. Either the hit is called homozygous inside the deletion, so it sits on the one copy left (the line and the carrier list call it hemizygous), or both calls carry read-backed phase (a `|` genotype with the same phase set, PS) on different copies;
+- **possibly biallelic** otherwise: the phase is unknown. Most short-read calls carry no phase, so this is the usual line. Step 05's checked table keeps no genotype column yet, and step 35's counts have no phase, so in a pipeline run a deletion pairs as possibly biallelic unless the hit is called homozygous inside it. A parent's genome (step 38, [family analysis](multi-sample.md#family-analysis-step-38-opt-in)) or long reads can settle it;
+- no line when read-backed phase puts both on the same copy (cis): the other copy is intact.
+
+Examples: an exon deletion of MUTYH plus a pathogenic MUTYH variant; three alpha-globin genes plus an HBA2 point variant; one SMN1 copy plus an SMN1 variant (short reads place few SMN1 variants reliably, so check with a targeted test); and the 1q21.1 deletion plus one of the two RBM8A low-expression alleles (c.-21G>A, rs139428292, or c.67+32G>C, rs201779890), the cause of TAR syndrome (thrombocytopenia with absent radius; Albers et al., Nat Genet 2012;44(4):435-439, PMID 22366785). On its own an RBM8A low-expression allele stays set aside. The [couple screen](39-couple.md) (step 39) applies the same pairing between partners: a hypomorph allele of the list in one partner and a full loss of function in the same gene in the other (a ClinVar pathogenic variant, a Delly deletion) is a shared gene with its risk for a child; the allele in both partners, or in one alone, stays set aside. In the secondary-findings card a called deletion counts as one of the two variants a recessive gene needs, unless read-backed phase puts it on the variant's copy. The line under the carrier list names the deletion callers it read and the ones it could not read. A deletion AnnotSV ranks below class 4 is not counted.
+
+The TTN rule (truncating variants only) and the LMNA rule (missense only with cardiac evidence) are printed beside the finding, not applied. It is still a list to review with a clinician.
 
 ---