Metrics

Category Metrics

---

Overview

Class for holding metrics on a single fingerprint. Note: this is distinct from {@link FingerprintingDetailMetrics} and {@link FingerprintingSummaryMetrics} in that it is calculated on a single fingerprint, and attempts to describe how likely that fingerprint is to have arisen from an actual sample as opposed to having artifacts such as contamination, or strong bias towards homozygous genotypes. Several tests use "expected" genotype distribution based on the allele frequency stored in the haplotype-database file that is used, and assuming Hardy-Weinberg equilibrium. Please see the FingerprintMetrics definitions http://broadinstitute.github.io/picard/picard-metric-definitions.html#FingerprintMetrics " + for a complete description of the metrics produced by this tool.

This table summarizes the values that are specific to this metric.

Metric	Summary
SAMPLE_ALIAS	The Sample alias taken from RG header or #CHROME line
SOURCE	The originating file (if available) for this fingerprint
INFO	Additional information about the fingerprint
HAPLOTYPES	Number of haplotypes examined
HAPLOTYPES_WITH_EVIDENCE	Number of haplotypes that had evidence in the source file
DEFINITE_GENOTYPES	Number of haplotypes that had enough evidence to make a definite genotype call (genotype LOD > Threshold
NUM_HOM_ALLELE1	Number of major allele homozygous calls
NUM_HOM_ALLELE2	Number of minor allele homozygous calls
NUM_HOM_ANY	Number of homozygous calls of either allele (might not be the sum of {@link #NUM_HOM_ALLELE1} and {@link #NUM_HOM_ALLELE1} due to rounding
NUM_HET	Number of heterozygous calls
EXPECTED_HOM_ALLELE1	Expected number of major allele homozygous calls
EXPECTED_HOM_ALLELE2	Expected number of major allele homozygous calls
EXPECTED_HET	Expected number of heterozygous calls
CHI_SQUARED_PVALUE	The Chi-squared pvalue of the observed counts vector relative to the expected counts, (3x2 table)
LOG10_CHI_SQUARED_PVALUE	The log10 of the Chi-squared pvalue
CROSS_ENTROPY_LOD	The categorical cross entropy of the counts of genotypes relative to expected (big is bad)
HET_CHI_SQUARED_PVALUE	The Chi-squared pvalue for the number of HETs and HOMs relative to the expected counts (2x2 table)
LOG10_HET_CHI_SQUARED_PVALUE	The log10 of the Chi-squared pvalue for the number of HETs and HOMs
HET_CROSS_ENTROPY_LOD	The categorical cross entropy of the counts of HETs and HOMs relative to the expected counts (big is bad)
HOM_CHI_SQUARED_PVALUE	The Chi-squared pvalue for {@link #NUM_HOM_ALLELE1} and {@link #NUM_HOM_ALLELE2} relative to the expected counts (2x2 table)
LOG10_HOM_CHI_SQUARED_PVALUE	The log10 of the Chi-squared pvalue for {@link #NUM_HOM_ALLELE1} and {@link #NUM_HOM_ALLELE2}
HOM_CROSS_ENTROPY_LOD	The categorical cross entropy of {@link #NUM_HOM_ALLELE1} and {@link #NUM_HOM_ALLELE2} relative to the expected counts (big is bad)
LOD_SELF_CHECK	The fingerprinting LOD score this sample gets when compared to itself (big is good)
DISCRIMINATORY_POWER	The difference in fingerprinting LOD between LOD_SELF_CHECK and the LOD score found when fingerprinting it against a random permutation of the probablity vectors (PLs) in each of the haplotypes. (big is good.)

---

Return to top

---

GATK version 4.6.0.0-33-gdffedfb built at Wed, 23 Oct 2024 21:44:48 -0400.