Clinical Variant Classification Predictor

Columns

• CHROM: Chromosome on which the variant is located.
• POS: Position of the variant on the chromosome.
• REF: Reference allele.
• ALT: Alternate allele.
• AF_ESP: Allele frequencies sourced from GO-ESP.
• AF_EXAC: Allele frequencies sourced from ExAC.
• AF_TGP: Allele frequencies sourced from the 1000 Genomes Project.
• CLNDISDB: Tag-value pairs providing the disease database name and identifier (e.g., OMIM:NNNNNN).
• CLNDISDBINCL: For included variants, tag-value pairs for disease database name and identifier.
• CLNDN: ClinVar's preferred disease name for the concept specified by disease identifiers in CLNDISDB.
• CLNDNINCL: For included variants, ClinVar's preferred disease name for the concept specified by disease identifiers in CLNDISDB.
• CLNHGVS: Top-level (primary assembly, alt, or patch) HGVS expression.
• CLNSIGINCL: Clinical significance for a haplotype or genotype that includes this variant, reported as pairs of VariationID:clinical significance.
• CLNVC: Type of variant (e.g., single_nucleotide_variant, Deletion).
• CLNVI: The variant's clinical sources reported as tag-value pairs of database and variant identifier.
• MC: Comma-separated list of molecular consequence in the format of Sequence Ontology ID|molecular_consequence.
• ORIGIN: Allele origin, which can include values like unknown, germline, somatic, inherited, paternal, maternal, de-novo, biparental, uniparental, not-tested, tested-inconclusive, or other.
• SSR: Variant Suspect Reason Codes, such as unspecified, Paralog, byEST, oldAlign, Para_EST, 1kg_failed, or other.
• CLASS: The target binary class, where '0' signifies no conflicting submissions and '1' signifies conflicting submissions.
• Allele: The variant allele used to calculate the consequence.
• Consequence: The type of consequence, as defined by Ensembl.
• IMPACT: The impact modifier for the consequence type (e.g., MODERATE, LOW).
• SYMBOL: Gene Name.
• Feature_type: Type of feature, typically Transcript, RegulatoryFeature, or MotifFeature.
• Feature: Ensembl stable ID of the feature.
• BIOTYPE: Biotype of the transcript or regulatory feature (e.g., protein_coding).
• EXON: The exon number (out of the total number).
• INTRON: The intron number (out of the total number).
• cDNA_position: Relative position of the base pair in the cDNA sequence.
• CDS_position: Relative position of the base pair in the coding sequence.
• Protein_position: Relative position of the amino acid in the protein.
• Amino_acids: Only provided if the variant affects the protein-coding sequence.
• Codons: The alternative codons with the variant base in upper case.
• DISTANCE: Shortest distance from the variant to the transcript.
• STRAND: Defined as + (forward) or - (reverse).
• BAM_EDIT: Indicates success or failure of editing using a BAM file (e.g., OK).
• SIFT: The SIFT prediction and/or score, typically given as prediction(score).
• PolyPhen: The PolyPhen prediction and/or score.
• MOTIF_NAME: The source and identifier of a transcription factor binding profile aligned at this position.
• MOTIF_POS: The relative position of the variation in the aligned TFBP.
• HIGH_INF_POS: A flag indicating if the variant falls in a high information position of a transcription factor binding profile (TFBP).
• MOTIF_SCORE_CHANGE: The difference in motif score between the reference and variant sequences for the TFBP.
• LoFtool: Loss of Function tolerance score for loss of function variants.
• CADD_PHRED: Phred-scaled CADD score.
• CADD_RAW: Score indicating the deleteriousness of variants.
• BLOSUM62: BLOSUM62 score.

Distribution

Usage

• Developing and testing machine learning models to predict conflicting clinical classifications of human genetic variants.
• Applying machine learning techniques to genomics research.
• Feature engineering to improve the assessment of variant classification consistency.
• Identifying single submission variants that may be prone to conflicting classifications in the future.

Coverage

License

Who Can Use It

• Genetics researchers and bioinformaticians looking to understand and mitigate classification discrepancies.
• Data scientists and machine learning engineers interested in applying predictive analytics to complex biological and healthcare data.
• Clinicians who interpret genetic variant classifications and need tools to identify potentially ambiguous cases.
• Academics and students exploring genomics and precision medicine applications using machine learning.

Dataset Name Suggestions

• ClinVar Variant Classification Conflict
• Predicting Conflicting Genetic Classifications
• Human Genome Variant Discrepancy
• Clinical Variant Classification Predictor

Attributes