SilencerDB

SilencerDB provides a curated and predicted catalogue of silencer elements to support analysis of sequence-specific genomic elements that negatively regulate gene transcription.

Key Features:

• Extensive Data Collection: Curated from over 2,300 published articles and containing 33,060 experimentally validated silencers and 5,045,547 predicted silencers.
• Advanced Prediction Model: Employs DeepSilencer, a deep convolutional neural network (CNN)-based model to classify and predict silencer elements.
• Structured Categorization: Implements a tree-structured class hierarchy to categorize silencers by species, organ, tissue, and cell line.
• Comprehensive Annotations: Annotates each silencer with information on the nearest gene and potential regulatory genes.

Scientific Applications:

• Transcriptional Regulation Studies: Supports investigation of mechanisms of transcriptional repression and modulation of gene expression during development and disease.
• Target Gene and Network Inference: Facilitates linking silencers to nearest genes and potential regulatory genes to study gene regulatory relationships.
• Comparative and Cell-type Analyses: Enables cross-species, organ-, tissue-, and cell line-specific analyses using the hierarchical categorization.

Methodology:

Entries were assembled via manual curation from published literature and potential silencers were predicted using the DeepSilencer deep convolutional neural network (CNN)-based model.