A novel procedure on next generation sequencing data analysis using text mining algorithm.

Area of Science:

• Bioinformatics
• Computational Biology
• Machine Learning

Background:

• Next-generation sequencing (NGS) generates vast biological and biomedical data, necessitating efficient data mining strategies for comparative and evolutionary studies.
• Topic modeling, a machine learning technique, is increasingly used for structuring large text corpora in data mining.

Purpose of the Study:

• To introduce a novel procedure for analyzing NGS data using topic modeling.
• To demonstrate the application of this procedure using Salmonella enterica strain data.
• To optimize the procedure through perplexity and convergence efficiency analysis.

Main Methods:

• A four-step procedure involving NGS data retrieval, preprocessing, topic modeling, and data mining using Latent Dirichlet Allocation (LDA).
• Application of LDA to Salmonella enterica NGS data.
• Evaluation of topic model performance using perplexity and Gibbs sampling convergence.

Main Results:

• LDA-derived topics accurately characterized the genetic diversity of the fliC gene across Salmonella serotypes.
• Hierarchical clustering and data matrix analysis of LDA outputs successfully classified Salmonella serotypes.
• The approach demonstrated potential for elucidating genetic information and identifying gene-phenotype relationships.

Conclusions:

• Topic modeling provides a novel method for NGS data analysis, enhancing genetic information extraction.
• This approach facilitates the identification of gene-phenotype relationships and biomarkers in the era of big biological data.