ChimerDB

ChimerDB integrates transcript sequences, literature, and next-generation sequencing analyses to detect and annotate fusion transcripts relevant to cancer.

Key Features:

• Enhanced Algorithm Sensitivity: An updated algorithm detects 2,699 fusion transcripts, including interchromosomal translocations and intrachromosomal deletions or inversions of large DNA segments.
• Integration of Next-Generation Sequencing Data: ChimerDB incorporates next-generation sequencing (NGS) results to improve identification of fusion events.
• Comprehensive Data Integration: Integrates mRNA and expressed sequence tag (EST) sequences from GenBank with manually collected literature data and entries from the Sanger Cancer Genome Project (CGP), OMIM, PubMed, and the Mitelman database.
• Classification of Fusion Events: Classifies fusion transcripts into genuine chromosome translocations and fusions between neighboring genes due to intergenic splicing and filters potential cloning artifacts.
• Deep Sequencing Data Analysis (ChimerSeq): ChimerSeq analyzes TCGA data from 10,565 patients using computational results from STAR-Fusion and FusionScan, compiling 65,945 fusion candidates with 21,106 predicted by multiple programs.
• Text Mining of Publications (ChimerPub): Applies deep learning–based text mining followed by manual curation to identify 1,257 fusion genes, including 777 cases supported by experimental evidence.
• Extensive Manual Annotations (ChimerKB): Contains 1,597 fusion genes with publication support, experimental evidence, and breakpoint information.
• Functional Significance Estimation Tools: Provides fusion structure annotations with protein domain information, gene expression plots comparing fusion-positive versus fusion-negative patients, and STRING network views to assess functional impact.

Scientific Applications:

• Biomarker and Therapeutic Target Discovery: Supports identification of fusion gene biomarkers and candidate drug targets in cancer by aggregating sequence, NGS, and literature evidence.
• Tumor Genomics and Tumorigenesis Studies: Enables exploration of genetic alterations associated with tumorigenesis through annotated fusion events and breakpoint information.
• Functional Assessment of Fusions: Facilitates assessment of functional significance by integrating domain annotations, expression comparisons, and protein interaction networks.

Methodology:

Integrates mRNA and EST sequences from GenBank and databases (Sanger CGP, OMIM, Mitelman) with NGS analyses; analyzes TCGA samples using STAR-Fusion and FusionScan; applies deep learning–based text mining of PubMed followed by manual curation; classifies and filters fusion candidates and aggregates cross-program predictions.