From a42d7dae63f4b9dd996fc9713677eb11f0bbed9a Mon Sep 17 00:00:00 2001
From: Ryan Amari <ramari16@gmail.com>
Date: Mon, 29 Jan 2024 14:58:04 -0500
Subject: [PATCH] Add logging to triage missing variant mask issue

---
 .../dbmi/avillach/hpds/data/genotype/VariantStore.java   | 2 +-
 .../hpds/processing/PatientVariantJoinHandler.java       | 9 +++++++--
 2 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/data/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/data/genotype/VariantStore.java b/data/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/data/genotype/VariantStore.java
index 822ea447..3efdc3df 100644
--- a/data/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/data/genotype/VariantStore.java
+++ b/data/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/data/genotype/VariantStore.java
@@ -142,7 +142,7 @@ public Optional<VariantMasks> getMasks(String variant, VariantBucketHolder<Varia
 			bucketCache.lastContig = contig;
 			bucketCache.lastChunkOffset = chrOffset;
 		}
-		return bucketCache.lastValue == null ? Optional.empty() : Optional.of(bucketCache.lastValue.get(variant));
+		return bucketCache.lastValue == null ? Optional.empty() : Optional.ofNullable(bucketCache.lastValue.get(variant));
 	}
 
 	public String[] getHeaders() {
diff --git a/processing/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/processing/PatientVariantJoinHandler.java b/processing/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/processing/PatientVariantJoinHandler.java
index 2538d6ba..c15c4b67 100644
--- a/processing/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/processing/PatientVariantJoinHandler.java
+++ b/processing/src/main/java/edu/harvard/hms/dbmi/avillach/hpds/processing/PatientVariantJoinHandler.java
@@ -1,5 +1,6 @@
 package edu.harvard.hms.dbmi.avillach.hpds.processing;
 
+import com.google.common.base.Joiner;
 import com.google.common.collect.Lists;
 import com.google.common.collect.Sets;
 import edu.harvard.hms.dbmi.avillach.hpds.data.genotype.VariantMasks;
@@ -70,8 +71,10 @@ public BigInteger getPatientIdsForIntersectionOfVariantSets(Set<Integer> patient
                 List<List<String>> variantBuckets = variantBucketPartitions.get(x);
                 variantBuckets.parallelStream().forEach(variantBucket -> {
                     VariantBucketHolder<VariantMasks> bucketCache = new VariantBucketHolder<>();
+                    List<String> missingVariants = new ArrayList<>();
                     variantBucket.forEach(variantSpec -> {
-                        variantService.getMasks(variantSpec, bucketCache).ifPresent(masks -> {
+                        Optional<VariantMasks> variantMask = variantService.getMasks(variantSpec, bucketCache);
+                        variantMask.ifPresentOrElse(masks -> {
                             BigInteger heteroMask = masks.heterozygousMask == null ? variantService.emptyBitmask() : masks.heterozygousMask;
                             BigInteger homoMask = masks.homozygousMask == null ? variantService.emptyBitmask() : masks.homozygousMask;
                             BigInteger orMasks = heteroMask.or(homoMask);
@@ -79,8 +82,10 @@ public BigInteger getPatientIdsForIntersectionOfVariantSets(Set<Integer> patient
                             synchronized(matchingPatients) {
                                 matchingPatients[0] = matchingPatients[0].or(andMasks);
                             }
-                        });
+                        }, () -> missingVariants.add(variantSpec));
                     });
+                    log.info(missingVariants.size() + "variant masks not found");
+                    log.info("Variants missing masks: " + Joiner.on(",").join( missingVariants.subList(0, 100)));
                 });
             }
             return matchingPatients[0];