Run

The run report name

Sample Name

Name of the sample

Reference

Name of the reference used in the analysis

Library Type

Library type

Variant Caller Version

Version of the plugin

Targeted regions

BED

Downloads the input targeted regions BED file (if any).

Hotspot regions

BED

Downloads the input hotspots BED file (if any).

Parameters Settings

Parameters File

Downloads a JSON text file of the TVC parameter values used on this run.

Note: You can edit this file and later upload it to set your custom parameters in subsequent runs. For an example of this type of file, see Example Torrent Variant Caller Parameter File .

Mapped Reads

BAM, BAI

Downloads t he BAM file (and its index) of mapped reads. This file is input to TVC.

Variant Calls

VCF.GZ,

VCF.GZ.TBI,

XLS

Downloads files of the variants calls:

VCF.GZ, VCF.GZ.TBI: Zipped VCF file and its tabix index file

XLS: T ab-separated values file

Open Variants Calls in IGV

IGV

Link to open the results variants in the Integrated Genomic Browser (IGV).

Deprecated Features

Classic

Opens the plugin results page in the previous format.

Position

The chromosome (or contig) name in the reference genome, and t he one-based position in the reference genome.

Ref

The reference base(s).

Variant

Variant allele base(s).

Var Freq

Frequency of the variant allele.

Quality

Phred-scored quality field.Larger values mean more certainty in the call.

Typically very large for reads strongly distinguishing variants (SNPs) with good depth; that is, under the model assumed, evidence is overwhelming for the variant or for the reference. Marginal values in this field can mean either the reads do not distinguish the variant well or there is insufficient depth to resolve, or the observed allele frequency is near the cutoff.Filters to compensate for the cases in which the model assumptions are not true are found in the INFO tags.

Computed by posterior probability that the sample variant allele frequency is greater than the min-allele-frequency specified for the variant type (if a variant), or posterior probability that the variant allele frequency is below this threshold (if a reference call).Posterior probability computed conditional on the reads observed, includes sampling variability.

SNP Single nucleotide polymorphism

IND Insertion

DEL Deletion

MNP Multiple nucleotide polymorphism

COMPLEX Complex block substitution

Hotspot if called only because of its entry in a hotspots file

Novel all others

Discrepancy between allele frequencies on the forward and reverse strands

Distance between predicted and observed signal at the allele locus. [RBI]

Distance between predicted and observed signal in the reference allele. [REFB]

Distance between predicted and observed signal in the variantallele. [VARB]

Phred-scaled mean log-likelihood difference between the prediction under reference and under the variant hypothesis. [MLLD]

Probability of sequence-specific error on the forward strand (reported only for deletion variants).

Probability of sequence-specific error on the reverse strand (reported only for deletion variants).

Basespace strand bias ( reported only for deletion variants ).

COVERAGE

Corresponds to the various counting VCF tags on each strand for alleles: FDP, FSAF, FSAR, FSRF, FSRR. Set by the TVC parameter min-cov-per-strand for the appropriate variant type, as well as parameter min-coverage .

FAIL

The filtering is based on the quality score (probability that the variant at the first location has an allele-frequency above the specified location, or for a reference call, that we can exclude the variant having an allele-frequency above the specified frequency), controlled by min-allele-frequency and min-variant-score .

Note : min-allele-frequency is the frequency at which the cutoff happens due to sampling variability, and the quality value for an observation at exactly this frequency is quite low.

HPLEN

Corresponds to the VCF tag HRUN. It is set by TVC parameter hp-max-length .This tag filters Insertion and deletion variants that happen in long runs.Defaults to 8.

PREDICTIONSHIFT

Corresponds to the VCF tag RBI, which measures the amount by which the predictions had to be distorted to match the measurements. A value such as 0.3=30% of the difference the variant would make to the read (e.g., a shift of 0.3 for a 1 base deletion, a shift of 0.6 for a 2-base deletion). Set by TVC parameter filter-unusual-predictions . Usual range is 0.2 - 0.3. Smaller values mean predictions are required to be more precise. Larger values mean that more shift is allowed.

REFERENCE

We generate a candidate variant at this location, but after considering the evidence, generate a 0/0 genotype call at the location. Because this clutters the VCF file with entries at locations that are not variants, there is a flag suppress-reference-calls that filters out candidates where we evaluate the genotype to be 0/0.

SSE

Corresponds to the VCF tags SSEN and SSEP, and means that at this location we predict a potential error due to the local sequence context, on either the positive or negative strand.Set by the two TVC internal filter controls sse-prob-threshold (0.2), min-ratio-reads-non-sse-strand (0.2). We filter on either of these scenarios:

• If both SSEN and SSEP are larger than the probability threshold set, or ...

• If only one is set and the other strand appears to be untrusted.

STDBIAS

Corresponds to the VCF tag STB.Filters out variants if the stand bias for the variant is too large. Set by TVC parameter strand bias (and the corresponding internal parameters snp-strand-bias , indel-strand-bias , and hotspot-strand-bias ).

STRINGENCY

Corresponds to the VCF tag MLLD, which is mean-log-likelihood difference per read for the variant. This is approximately phred-scaled, so that MLLD=10 = 10% error rate per read, MLLD=20 = 1% error rate per read. Variants are filtered by the TVC parameter data-quality-stringency .Smaller values imply worse performance per read.