BioSeq-Diabolo

BioSeq-Diabolo performs biological sequence similarity analysis using Biological Language Semantics (BLS) to assess structure, function, and disease associations across DNA, RNA, protein, and disease-related sequences.

Key Features:

• Biological Language Semantics (BLS): applies NLP-inspired semantic modeling to biological sequences.
• NLP-derived semantics methods: implements 27 semantics analysis methods derived from natural language processing.
• Embeddings and task identification: accepts embeddings of biological sequence data and performs intelligent task identification.
• Learning to Rank (LTR) integration: integrates multiple sequence similarity measures using supervised Learning to Rank to evaluate method performance and recommend methods.
• Sequence types and low-similarity challenges: handles DNA, RNA, protein, and disease-related sequences and addresses low sequence similarity challenges such as remote homology.
• Analytical outputs: enables detection of protein remote homology, identification of circRNA-disease associations, and annotation of protein functions.

Scientific Applications:

• Protein remote homology detection: identifies remote homologs among protein sequences.
• circRNA-disease association prediction: identifies associations between circRNA and diseases.
• Protein function annotation: annotates protein functions based on sequence similarity semantics.
• Genomics and proteomics research and disease studies: provides similarity-based insights to support genomics and proteomics investigations and inform disease diagnosis and treatment research.

Methodology:

Uses Biological Language Semantics with 27 NLP-derived semantics analysis methods on sequence embeddings, performs intelligent task identification, and applies supervised Learning to Rank (LTR) to integrate sequence similarity measures and evaluate and recommend methods.