tb-profiler

tb-profiler predicts drug resistance in Mycobacterium tuberculosis from genomic data to generate drug-susceptibility profiles and strain-type information.

Key Features:

• Mutation library: Leverages a curated library of 1,325 mutations predictive of drug resistance for 15 anti-tuberculosis drugs.
• Validation dataset: The mutation library was compiled and validated using genomic–phenotypic data from 792 strains.
• Drug validation: Predictive calls have been validated for 11 anti-tuberculosis drugs.
• Input formats: Accepts raw sequencing reads, alignments, or variant data as input.
• Outputs: Produces detailed drug-resistance (DR) profiles and strain-type information.
• Reported performance: Diagnostic accuracy has been reported to exceed some existing commercial diagnostics and alternative databases.

Scientific Applications:

• Sequence-based drug-susceptibility testing: Enables drug-susceptibility calls directly from genomic sequences.
• Clinical decision support: Provides drug-resistance predictions to inform treatment decisions.
• Research and surveillance: Supports tuberculosis research, strain typing, and management by translating genomic data into actionable DR profiles.

Methodology:

Analyzes raw sequencing reads, alignments, or variant data against a curated library of 1,325 mutations to generate drug-resistance profiles and strain-type information; the mutation library was compiled and validated using genomic–phenotypic data from 792 strains and validated for 11 drugs.