From d8a761bbbd631a6c32f78549b250072b0c476886 Mon Sep 17 00:00:00 2001
From: rpoplin <rpoplin@348d0f76-0448-11de-a6fe-93d51630548a>
Date: Fri, 6 May 2011 17:47:58 +0000
Subject: [PATCH] Warn the user if trying to train with too few variants

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5776 348d0f76-0448-11de-a6fe-93d51630548a
---
 .../variantrecalibration/VariantRecalibratorEngine.java        | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/java/src/org/broadinstitute/sting/gatk/walkers/variantrecalibration/VariantRecalibratorEngine.java b/java/src/org/broadinstitute/sting/gatk/walkers/variantrecalibration/VariantRecalibratorEngine.java
index e87048a13..21b7f1bdf 100755
--- a/java/src/org/broadinstitute/sting/gatk/walkers/variantrecalibration/VariantRecalibratorEngine.java
+++ b/java/src/org/broadinstitute/sting/gatk/walkers/variantrecalibration/VariantRecalibratorEngine.java
@@ -32,6 +32,9 @@ public class VariantRecalibratorEngine {
     }
 
     public GaussianMixtureModel generateModel( final List<VariantDatum> data ) {
+        if( data.size() < 1500 ) {
+            logger.warn("WARNING: Training with very few variant sites! Please check the model reporting PDF to ensure the quality of the model is reliable.");
+        }
         final GaussianMixtureModel model = new GaussianMixtureModel( VRAC.MAX_GAUSSIANS, data.get(0).annotations.length, VRAC.SHRINKAGE, VRAC.DIRICHLET_PARAMETER, VRAC.PRIOR_COUNTS );
         variationalBayesExpectationMaximization( model, data );
         return model;