iLoc-miRNA

iLoc-miRNA predicts the subcellular localization of microRNAs (miRNAs) in Homo sapiens to distinguish extracellular versus intracellular locations for studies of regulatory function and cell communication.

Key Features:

• Deep Learning Architecture: Employs bidirectional long short-term memory (BiLSTM) networks combined with a multi-head self-attention mechanism to capture complex sequence patterns.
• Data Representation: Uses one-hot encoding with post-padding to represent full-length miRNA sequences.
• Functionally Similar Location Categorization: Applies a novel data partitioning strategy to categorize functionally similar subcellular locations in Homo sapiens.
• High Selectivity: Distinguishes extracellular and intracellular miRNAs with high selectivity.
• Motif Analysis: Performs motif analyses to investigate sequence signals associated with subcellular localization.

Scientific Applications:

• Regulation Activity Studies: Enables analysis of miRNA regulatory roles by identifying their extracellular or intracellular localization.
• Cell-to-Cell Communication Research: Facilitates study of extracellular miRNAs involved in intercellular communication.
• Stress Response Analysis: Supports investigation of intracellular miRNA distribution changes in response to cellular stress and environmental stimuli.

Methodology:

Computational methods include one-hot encoding with post-padding for sequence representation, bidirectional LSTM networks integrated with multi-head self-attention for sequence modeling, a novel data partitioning strategy to categorize functionally similar locations, and motif analysis.