feat: script fix_glnexus.py prepare GLNexus output for noodles (#356)

misc/fix_glnexus.py

@@ -0,0 +1,64 @@
+#!/usr/bin/env python3
+"""
+Fix GLNexus output for input to `mehari annotate seqvars`.
+
+First, note that multiallelic sites must be corrected by a call to
+`bcftools norm` already as an input to this script.
+
+What we will do is splitting the `FORMAT/RNC` field with a comma
+as noodles-vcf expects a list of characters for rather than a
+string of length 2 for `##FORMAT=<ID=RNC,Number=2,Type=Character>`.
+"""
+
+from pathlib import Path
+import sys
+from typing import Annotated
+
+import typer
+import vcfpy
+
+
+def main(
+    path_in: Annotated[Path, typer.Option(help="Path to input VCF")],
+    path_out: Annotated[Path, typer.Option(help="Path to output VCF")],
+    quiet: Annotated[bool, typer.Option(help="Disable verbose output")] = False,
+):
+    """
+    Fix GLNexus output to be compatible with `noodles-vcf`.
+
+    cf: https://github.com/varfish-org/mehari/issues/356
+    """
+    if not quiet:
+        print(f"Opening input file {path_in}", file=sys.stderr)
+    reader = vcfpy.Reader.from_path(path_in)
+    header = reader.header.copy()
+    for line in header.lines:
+        if line.key == "FORMAT" and line.mapping.get("ID") == "GQ":
+            line.mapping["Number"] = 1
+            line.mapping["Type"] = "Integer"
+    if not quiet:
+        print(f"Opening output file {path_out}", file=sys.stderr)
+    writer = vcfpy.Writer.from_path(path_out, header)
+    if not quiet:
+        print("Processing records...", file=sys.stderr)
+    with reader, writer:
+        for idx, record in enumerate(reader):
+            if idx % 10_000 == 0:
+                print(
+                    f"  at {idx} records {record.CHROM}:{record.POS}", file=sys.stderr
+                )
+            for call in record.calls:
+                if "RNC" in call.data:
+                    if (
+                        isinstance(call.data["RNC"], list)
+                        and len(call.data["RNC"]) == 1
+                    ):
+                        call.data["RNC"] = list(call.data["RNC"][0])
+            writer.write_record(record)
+    if not quiet:
+        print("... done", file=sys.stderr)
+        print("All done. Have a nice day!", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    typer.run(main)

src/annotate/seqvars/csq.rs

@@ -1056,15 +1056,15 @@ mod test {
     // Compare to SnpEff annotated variants for OPA1, touching special cases.
     #[test]
     fn annotate_opa1_hand_picked_vars() -> Result<(), anyhow::Error> {
-        annotate_opa1_vars("tests/data/annotate/vars/opa1.hand_picked.tsv", true)
+        annotate_opa1_vars("tests/data/annotate/seqvars/opa1.hand_picked.tsv", true)
     }
 
     // Compare to SnpEff annotated ClinVar variants for OPA1 (slow).
     #[ignore]
     #[test]
     fn annotate_opa1_clinvar_vars_snpeff() -> Result<(), anyhow::Error> {
         annotate_opa1_vars(
-            "tests/data/annotate/vars/clinvar.excerpt.snpeff.opa1.tsv",
+            "tests/data/annotate/seqvars/clinvar.excerpt.snpeff.opa1.tsv",
             true,
         )
     }
@@ -1074,7 +1074,7 @@ mod test {
     #[test]
     fn annotate_opa1_clinvar_vars_vep() -> Result<(), anyhow::Error> {
         annotate_opa1_vars(
-            "tests/data/annotate/vars/clinvar.excerpt.vep.opa1.tsv",
+            "tests/data/annotate/seqvars/clinvar.excerpt.vep.opa1.tsv",
             true,
         )
     }
@@ -1095,15 +1095,15 @@ mod test {
     // Compare to SnpEff annotated variants for BRCA1, touching special cases.
     #[test]
     fn annotate_brca1_hand_picked_vars() -> Result<(), anyhow::Error> {
-        annotate_brca1_vars("tests/data/annotate/vars/brca1.hand_picked.tsv", true)
+        annotate_brca1_vars("tests/data/annotate/seqvars/brca1.hand_picked.tsv", true)
     }
 
     // Compare to SnpEff annotated ClinVar variants for BRCA1 (slow).
     #[ignore]
     #[test]
     fn annotate_brca1_clinvar_vars_snpeff() -> Result<(), anyhow::Error> {
         annotate_brca1_vars(
-            "tests/data/annotate/vars/clinvar.excerpt.snpeff.brca1.tsv",
+            "tests/data/annotate/seqvars/clinvar.excerpt.snpeff.brca1.tsv",
             true,
         )
     }
@@ -1113,7 +1113,7 @@ mod test {
     #[test]
     fn annotate_brca1_clinvar_vars_vep() -> Result<(), anyhow::Error> {
         annotate_brca1_vars(
-            "tests/data/annotate/vars/clinvar.excerpt.vep.brca1.tsv",
+            "tests/data/annotate/seqvars/clinvar.excerpt.vep.brca1.tsv",
             true,
         )
     }

src/annotate/seqvars/mod.rs

@@ -654,7 +654,7 @@ impl VarFishSeqvarTsvWriter {
         P: AsRef<Path>,
     {
         Self {
-            inner: if p.as_ref().extension().unwrap() == "gz" {
+            inner: if p.as_ref().extension().unwrap_or_default() == "gz" {
                 Box::new(GzEncoder::new(
                     File::create(p).unwrap(),
                     Compression::default(),
@@ -1501,7 +1501,17 @@ fn run_with_writer(writer: &mut dyn AnnotatedVcfWriter, args: &Args) -> Result<(
         if let Some(record) = records.next() {
             let mut vcf_record = record?;
 
-            // TODO: ignores all but the first alternative allele!
+            // We currently can only process records with one alternate allele.
+            if vcf_record.alternate_bases().len() != 1 {
+                tracing::error!(
+                    "Found record with more than one alternate allele.  This is currently not supported. \
+                    Please use `bcftools norm` to split multi-allelic records.  Record: {:?}",
+                    &vcf_record
+                );
+                anyhow::bail!("multi-allelic records not supported");
+            }
+
+            // Get first alternate allele record.
             let vcf_var = keys::Var::from_vcf_allele(&vcf_record, 0);
 
             // Skip records with a deletion as alternative allele.
@@ -1769,19 +1779,20 @@ mod test {
             transcript_source: TranscriptSource::Both,
             transcript_picking: false,
             path_db: String::from("tests/data/annotate/db"),
-            path_input_vcf: String::from("tests/data/db/create/badly_formed_vcf_entry.vcf"),
+            path_input_vcf: String::from("tests/data/annotate/seqvars/badly_formed_vcf_entry.vcf"),
             output: PathOutput {
                 path_output_vcf: None,
                 path_output_tsv: Some(path_out.into_os_string().into_string().unwrap()),
             },
             max_var_count: None,
-            path_input_ped: String::from("tests/data/db/create/badly_formed_vcf_entry.ped"),
+            path_input_ped: String::from("tests/data/annotate/seqvars/badly_formed_vcf_entry.ped"),
         };
 
         run(&args_common, &args)?;
 
         let actual = std::fs::read_to_string(args.output.path_output_tsv.unwrap())?;
-        let expected = std::fs::read_to_string("tests/data/db/create/badly_formed_vcf_entry.tsv")?;
+        let expected =
+            std::fs::read_to_string("tests/data/annotate/seqvars/badly_formed_vcf_entry.tsv")?;
         assert_eq!(&expected, &actual);
 
         Ok(())
@@ -1805,19 +1816,59 @@ mod test {
             transcript_source: TranscriptSource::Both,
             transcript_picking: false,
             path_db: String::from("tests/data/annotate/db"),
-            path_input_vcf: String::from("tests/data/db/create/mitochondrial_variants.vcf"),
+            path_input_vcf: String::from("tests/data/annotate/seqvars/mitochondrial_variants.vcf"),
+            output: PathOutput {
+                path_output_vcf: None,
+                path_output_tsv: Some(path_out.into_os_string().into_string().unwrap()),
+            },
+            max_var_count: None,
+            path_input_ped: String::from("tests/data/annotate/seqvars/mitochondrial_variants.ped"),
+        };
+
+        run(&args_common, &args)?;
+
+        let actual = std::fs::read_to_string(args.output.path_output_tsv.unwrap())?;
+        let expected =
+            std::fs::read_to_string("tests/data/annotate/seqvars/mitochondrial_variants.tsv")?;
+        assert_eq!(&expected, &actual);
+
+        Ok(())
+    }
+
+    /// Test some variants originating from Clair3 on ONT data merged via GLNexus.
+    ///
+    /// Note that we currently re-use the GRCh37 databases in
+    /// `tests/data/annotate/db/grch37` for GRCh38 via symlink.  As the below
+    /// only is a smoke test, this is sufficient.  However, for other tests,
+    /// this will pose a problem.
+    #[test]
+    fn test_clair3_glnexus_variants() -> Result<(), anyhow::Error> {
+        let temp = TempDir::default();
+        let path_out = temp.join("output.tsv");
+
+        let args_common = crate::common::Args {
+            verbose: Verbosity::new(0, 1),
+        };
+        let args = Args {
+            genome_release: None,
+            report_all_transcripts: true,
+            transcript_source: TranscriptSource::Both,
+            transcript_picking: false,
+            path_db: String::from("tests/data/annotate/db"),
+            path_input_vcf: String::from("tests/data/annotate/seqvars/clair3-glnexus-min.vcf"),
             output: PathOutput {
                 path_output_vcf: None,
                 path_output_tsv: Some(path_out.into_os_string().into_string().unwrap()),
             },
             max_var_count: None,
-            path_input_ped: String::from("tests/data/db/create/mitochondrial_variants.ped"),
+            path_input_ped: String::from("tests/data/annotate/seqvars/clair3-glnexus-min.ped"),
         };
 
         run(&args_common, &args)?;
 
         let actual = std::fs::read_to_string(args.output.path_output_tsv.unwrap())?;
-        let expected = std::fs::read_to_string("tests/data/db/create/mitochondrial_variants.tsv")?;
+        let expected =
+            std::fs::read_to_string("tests/data/annotate/seqvars/clair3-glnexus-min.tsv")?;
         assert_eq!(&expected, &actual);
 
         Ok(())