Tumor mutation analysis

The Tumor mutation mnalysis (TMA) report builder filters, annotates, and summarizes variant data from tumor samples or from paired tumor versus normal sample pairs. TMA then adds annotation and applies user-defined filters to each variant. Variant annotation includes Variant Effect Prediction (VEP) impact and score, COSMIC annotation, relevant FDA approved drugs, and other publicly available data.

Tumor Mutation Analysis (TMA) in Sequence Miner

Example use case

The user has normal and tumor matched sample pairs and wishes to identify all likely somatic variants and to determine which variants are actionable and occur in known cancer genes.

Description of the algorithm

The processing pipeline utilizes the calls from each algorithm. The TMA applies user-defined filters to these lists of variants and adds several annotations.

Interpreting the output

The output can be displayed in Table view or Record view. The Table view of the output contains more than 90 columns. In general, these columns fall into three major categories:

• Basic information (Chrom, Pos, Reference, Call, PN)

• Annotations generated by the variant calling algorithms

• Annotations from public data sources including COSMIC and GO annotations, actionable mutations and pathways, and clinical annotations

The Record view displays the variants in a list format and summarizes the most relevant information about the variant and key annotations:

• Row 1: Chromosome position, gene, links to databases, transcript ID

• Row 2: Nucleotide change, amino acid change and position, VEP maximum consequence, VEP maximum impact, maximum allele frequency, conservation score (maximum from SIFT and Polyphen-2)

• Row 3: Callers for the variant, reads/depth in the tumor and germline samples

• Row 4: COSMIC annotations

Record view

To toggle from Table view to Record view, click the list icon.

Column descriptions

Basic columns and descriptions

Group	Column	Description
Basic	Call	Sequence (variant) called, based on the reference sequence at the designated position
	Chrom	Variant chromosome location
	PN	Patient number (identifier)
	Pos	Variant start bp position
	Reference	Nucleotide from the reference build at the base pair position in the Pos column

COMM columns and descriptions

Group	Column	Description
COMM	CLINICAL_SIGNIFICANCE	The clinical significance (e.g., pathogenic, benign, unknown significance, drug-response, risk factor, etc.) of the variant as annotated (commented) by users; if the same variant has several comments, this cell will contain a set of values
	MODE_OF_INHERITANCE	The user-annotated (commented) mode of inheritance of the variant; if the same variant has been commented several times, this cell will contain a set of values
	TEXT	The description (comment) component for the user annotation of the variant

COSMIC columns and descriptions

Group	Column	Description
COSMIC	count_lociInGene	Count of COSMIC variants in this gene
	count_primarysitesInGene	Number of primary site annotations in this gene
	count_vars	Count of occurences of variants in COSMIC
	count_varsInGene	Count of occurences of variants in this gene in COSMIC
	HGVSp	Variant HGVSp notation using the preferred transcript; present only if the gene has a preferred transcript and the variant is found in this transcript
	IDs	All COSMIC IDs for the variant
	primarysites	List of primary site annotations for this variant
	primarysitesInGene	List of primary site annotations in this gene
	pscount	Number of primary site annotations for this variant

Drug columns and descriptions

Group	Column	Description
Drug	Gene	Cancer drugs targeting this gene
	Gene_ApprovalStatus	Approval status of the cancer drug
	Mechanism_of_Action	Mechanism of action of the cancer drug
	Pathway	Cancer drugs targeting a pathway of which this gene is a member
	Pathway_ApprovalStatus	Approval status of the cancer drug
	SourcePathway	The pathway on which the cancer drug acts

Gene columns and descriptions

Group	Column	Description
Gene	ActionableDrugClasses	Drug classes associated with the gene encoding the variant
	ActionableMutations	Clinically actionable mutations within genes
	ActionablePathways	Clinically actionable pathways associated with genes
	ActionableTumors	Clinically actionable tumors associated with genes
	Aliases	List of gene aliases that correspond to the GENE_symbol
	listcount	A count of the number of Gene_lists
	lists	Gene panels or Gene lists
	Paralogs	The paralogs of a given gene
	symbol	Based on HGNC when it exists, otherwise it is the Ensembl internal alias

Germ columns and descriptions

Group	Column	Description
Germ	depth	The number of reads used in evaluating the corresponding call
	depth_hq	The number of high-quality reads used in evaluating the corresponding call
	reads	The number of reads containing the variant call
	reads_hq	The number of high quality reads containing the variant call

GO columns and descriptions

Group	Column	Description
GO	Descriptions	Gene Ontology category descriptions
	IDs	Gene Ontology identifiers

KNOWN columns and descriptions

Group	Column	Description
KNOWN	dbSource	Variants known in databases like HGMD, OMIM and ClinVar
	Gene_diseases	Diseases known to be associated with the gene as annotated by HGMD, ClinVar, and OMIM
	lis_dbSourceMaxClinImpact	A comma-delimited list of the KNOWN_dbsource and the known clinical impact of the gene (pathogenic/nonpathogenic, etc.)
	lis_disease	A comma-delimited list of associated KNOWN_diseases
	MaxClinImpact	The known clinical impact of the gene (pathogenic/nonpathogenic, etc.)
	var_diseases	Diseases known to be associated with the variant as annotated by HGMD, ClinVar, and OMIM

Mutect2 columns and descriptions

Group	Column	Description
Mutect2	failure_reasons	The reason MuTect judges a variant to be low confidence (REJECT)
	FORMAT	Mutect2 Genotype Format Fields
	INFO:	Mutect2 Genotype Format Info
	judgement	MuTect judgement of site as somatic or not (KEEP or REJECT (not enough evidence or artifact))
	NORMAL	Genotype data from the normal sample
	Normal_LOD	Log odds ratio (LOD) score for the normal sample
	TUMOR	Log odds ratio (LOD) score for the normal sample
	Tumor_LOD	Log odds ratio (LOD) score for tumor sample

OMIM columns and descriptions

Group	Column	Description
OMIM	Descriptions	OMIM disease descriptions for the gene
	IDs	The OMIM ID of the gene

Tumor columns and descriptions

Group	Column	Description
Tumor		High-quality reads are those with MAPQ >= 20 (MAPQ = -10*log10 (probability mapping position is incorrect))
	depth_all	The number of reads used in evaluating the corresponding call
	depth_hq	The number of high-quality reads used in evaluating the corresponding call
	reads_all	The number of reads containing the variant call
	reads_hq	The number of high quality reads containing the variant call

varscan2 columns and descriptions

Group	Column	Description
varscan2	FILTER	Judgement of the variant
	FORMAT	VarScan2 Genotype Format Fields
	HighConf	VarScan2 further classify the somatic mutations as high-confidence or low-confidence
	INFO	VarScan2 Genotype Format Info
	NORMAL	Genotype data from the normal sample
	somatic_p_value	Variant p-value for Somatic/LOH events
	somatic_p_score	Score based on the p-value
	somatic_status	Somatic status call (Germline, Somatic, LOH, or Unknown)
	TUMOR	Genotype data from the tumor sample
	variant_p_value	Variant p-value for germline events

Variant columns and descriptions

Group	Column	Description
Variant	Caller	Variant caller used to judge the variant
	Caller_type	Variant caller type: unpaired or paired
	PASS_count	Number of variant callers that approve or “pass” the variant
	Present_count	Number of variant callers that judge the variant

Venn A columns and descriptions

Group	Column	Description
Venn A	In”Variant caller”	A Boolean value if the variant has been rejected (0) or passed by the variant caller

Venn B columns and descriptions

Group	Column	Description
Venn B	inGenelist	A Boolean value if the variant is absent (0) or present (1) in a cancer gene list.
	WithPrimeSite	Variant occurs in a tumor primary site as reported by COSMIC (1) or not (0)

VEP columns and descriptions

Group	Column	Description
VEP	Amino_Acids	The amino acid change given as reference AA/variant AA; if the variation does not affect the protein-coding sequence, then “.”
	Biotype	Biological class of transcript or regulatory feature
	CDS_position:	Base pair position of the variant in the coding sequence; a value is given for each transcript
	HGVSp	Variant notation with the transcript and amino acid change; there is an entry for each transcript
	HGVSp_simple	Variant notation with the amino acid change; there is an entry for each transcript
	Max_Af	Maximum reported allele frequency across the population surveys from 1000GP3, EVS, EXAC, Kyoto, GONL, and Iceland (Variant View-Frequencies)
	max_consequence	Variant classes (high, moderate, low, and/or lowest impact on the gene product)
	Max_Impact	Classification of the level of severity of the transcript consequence type assigned by VEP
	Max_Score	Maximum score for the variant as observed in dbNSFP [Score=max ((1-Sift_score), Polyphen2_HDIV_score, Polyphen2_HVAR_score)]
	Protein_Position	Position of the amino acid in the protein sequence (only if the variant falls within a coding sequence); a value is given for each corresponding transcript specified in the CDS position field
	Refgene	The accession number from NCBI of the affected transcripts
	Transcript_count:	Number of different transcripts in which the variant is found

Other columns

Group	Column	Description
Other columns	Annotation_Gene	Cancer annotation for this gene
	CallRatio	Proportion of reads containing the variant call; expected to be approximately 0.5 for heterozygous calls and close to 1 for homozygous calls
	candidate_gene	Candidate genes entered by the user
	Category	Denotes if variant is present in an oncogene or tumor suppressor
	dbNSFP_Interpro_domain	The annotation from the dbNSF database, the domain of protein
	DIAG_ACMGcat	Categorization of the sequence variants according to the ACMG scheme
	ReadDiff	Difference in call ratio between tumor and normal samples
	Entrez-ID	Entrez gene nomenclature
	ResistancexSensitivity_Variants	Resistant to sensitivity information related to the variant
	SubjectId	Patient ID

Perspective views

The Default view perspective lists all variants depending on the variantFilter criteria selected in the input parameters. Additional perspectives focus on subsets of the columns in the default view.

Perspectives

Perspective	Description
Candidate genes	Somatic variants present in the user’s candidate genes report file (if selected from the user in the input parameters)
Default view	Lists all variants depending on the variantFilter criteria selected in the input parameters, e.g., if the user selects “Filtered variants”, this perspective will list the same number of variants shown in the “Filtered” perspective
Filtered	Somatic variants that meet the quality criteria (present in at least 2 high-quality reads and ratio of hq reads to depth in tumor is greater than 0.01); high-quality reads are those with MAPQ ≥ 20 (MAPQ = -10*log10 (probability mapping position is incorrect))
Filtered cancer genes	Somatic variants that meet the quality criteria (see Filtered perspective above) and are present in cancer-associated genes (from commercial gene panels); the list of commercial gene panels is listed in the ref/cancer/cancer_commercial.map file
Filtered candidate genes	Somatic variants that meet the quality criteria (see Filtered perspective above) and are present in the user’s candidate genes report file (if selected from the user in the input parameters)

The user can then filter on additional annotation (e.g., COSMIC annotation, known associated gene disease, and actionable mutations/pathways) to further narrow the list of candidate genes and gene variants. These annotations can be found in the right-hand pane, which lists all of the columns, grouped by category.