Additional Metrics Information
The file name for the Read Collapsing Metrics JSON file is <SampleID>_metrics.json. This JSON file contains the following read collapsing metrics:
Metric Name |
Description |
Calculation |
---|---|---|
MeanFamilyDepth |
Average number of fragments (paired reads) in a family before UMI collapsing. |
SUM(Reads in each family)/ the number of families after correction, collapsing, and filtering on supporting reads. |
PercentFamiliesCollapsed |
The percentage of the total number of families that are also collapsed. |
100 * FamiliesWritten / (FamiliesCreated - FamiliesContextuallyCorrected - FamiliesShifted - DuplexFamilies) |
PercentDuplexFamilies |
The percentage of the total number of families that are duplex families. |
100 * DuplexFamilies / (FamiliesCreated - FamiliesContextuallyCorrected - FamiliesShifted - DuplexFamilies) |
PercentReadsWithUsableUmis |
The percentage of reads with usable UMIs. |
100 * (ReadsOnTarget- FilteredOutReads - ReadsDiscardedInvalidUmis) / (ReadsOnTarget - FilteredOutReads) |
CvUmiCoverage |
The coefficient of variation of the number of families associated with each UMI. |
Math.Sqrt(UmiFamilyCounts.Average(v => MAvg.Pow(v - avg, 2))) / UmiFamilyCounts.Average() |
ReadsOnTarget |
The number of raw reads that overlap with manifest intervals. |
Incremented by one for each read. |
FilteredOutReads |
The number of raw reads that are discarded before creating families, including those unpaired, mate unmapped, without cigar string, or low map quality. |
Incremented by one for each read. |
SVCandidateReads |
The number of raw reads that are considered as structural variation candidates, including fragment that exceeds max size specified or mapped onto the different chromosomes. |
Incremented by one for each read. |
FamiliesCreated |
The total number of families created. |
Incremented by one for each family |
FamiliesContextuallyCorrected |
The number of families that are merged into other families based on contextual UMI correction. |
Incremented by one for each family. |
FamiliesShifted |
The number of families merged into other families based on proximity with matching UMIs. |
Incremented by one for each family. |
FamiliesWritten |
The number of families written out in the form of a condensed read pair. |
Incremented by one for each family. |
DuplexFamilies |
The number of families that meet the criteria for cross-collapsing. |
Incremented by one for each family. |
FamiliesDiscarded |
The number of families that are discarded during correction. |
Incremented by one for each family. |
FamiliesDiscardedInvalidUmis |
The number of invalid families that are discarded, including uncorrectable UMIs. |
Incremented by one for each family. |
ReadsDiscardedInvalidUmis |
The number of invalid reads that are discarded, including uncorrectable UMIs. |
|
ReadsInDiscardedFamilies |
The number of reads that are thrown away when discarding families. |
Incremented by total reads in each discarded family. |
ReadsWithValidOrCorrectedUmi |
The number of reads with a particular UMI after UMI correction, prior to collapsing and filtering on supporting reads. |
Incremented by two for each read. |
FamiliesWithValidOrCorrectedUmi |
The number of families with a particular UMI after correction, collapsing but before filtering on supporting reads. |
Incremented by one for each family. |
NumSupportingFragments |
An array of counts for how many families have this many reads supporting it. Calculated after UMI correction, collapsing, and filtering on supporting reads. |
Incremented by one for each read per family. |
FragmentSizeHistogram |
The histogram of fragment size of condensed (consensus) families. |
Incremented by one for each read |
The file name for the Error Rate Metrics JSON file is <SampleID>_noiseAF.json. This JSON file contains the following metrics:
Name |
Description |
Mathematical definition |
---|---|---|
Total_DP |
Total Depth of Fragment Coverage |
Sum of depth of fragment coverage (nA+nC+nG+nT+nIndel) |
Total_AD |
Total Depth of Alternative Allele |
Sum of alternative allele. Number of alternative allele = Min (total counts different from the reference base, total counts same as the reference base). Note: when a patient whose genotype is the homozygous variant, total alternative depth is total counts same as the Reference base. |
NoiseAF |
Noise Allele Frequency/Error Rate |
Average of all allele frequency values between 0 and 0.05 excluding loci overlapping with known germline mutations from 1000 genome data Fnset. NoiseAF= total_AD/total_DP |
MTC |
Median Target Coverage across all base positions |
Median of depth of fragment coverage (nA+nC+nG+nT+nIndel) across all base positions. |
CV_MTC |
Percentage Coefficient of Variation of Target Coverage across all base positions |
100 * median absolute deviation of depth of fragment coverage (nA+nC+nG+nT+nIndel) / Median of depth of fragment coverage (nA+nC+nG+nT+nIndel). |
PCT_1000X |
Parameter insensitive to input manifest |
(Number of positions with depth of fragment coverage >1000)/(Total number of positions) *100% |
PCT_1500X |
Parameter insensitive to input manifest |
(Number of positions with depth of fragment coverage >1500)/(Total number of positions) *100% |
PCT_2000X |
Parameter insensitive to input manifest |
(Number of positions with depth of fragment coverage >2000)/(Total number of positions) *100% |