Quikr

Quikr performs rapid taxonomic profiling of microbial communities from 16S rRNA amplicon sequencing by reconstructing community composition using a quadratic, K-mer-based iterative algorithm informed by compressed sensing.

Key Features:

• Optimization-Based Approach: Uses an optimization framework inspired by compressive sensing to compute vectors representing taxonomic assignments and their proportions within a sample.
• K-mer-based Quadratic Iterative Reconstruction: Employs a quadratic, K-mer-based iterative reconstruction technique to infer community composition from sequence data.
• Speed and Efficiency: Processes large datasets orders of magnitude faster than read-by-read methods such as the Ribosomal Database Project's Naïve Bayesian Classifier.
• Lower Error Rates: Demonstrates reduced taxonomic assignment error in both simulated and actual biological datasets.
• Chimera Resilience: Is reported to be unaffected by the presence of chimeric sequences, removing the need for chimera filtering steps.
• Multi-language Implementations: Provides computational implementations in MATLAB, Octave, Python, and C.

Scientific Applications:

• OTU profiling from 16S rRNA amplicons: Infers operational taxonomic unit composition directly from 16S rRNA amplicon sequencing data.
• Beta-diversity and ecological assessment: Facilitates analysis of beta-diversity metrics and other community-level ecological assessments by supplying taxonomic abundance profiles.
• Comparative studies across sample sets: Supports comparative metagenomic analyses of environmental and health-related samples by enabling rapid taxonomic comparison.

Methodology:

Performs quadratic, K-mer-based iterative reconstruction using sparsity-promoting ideas from compressed sensing to compute taxonomic abundance vectors and reconstruct bacterial community composition from 454-like pyrosequencing and 16S amplicon data without explicit chimera filtering.