Cnngeno

Cnngeno predicts genotypes of structural variations from next-generation sequencing data using convolutional neural network–based deep learning models.

Key Features:

• CNN-Based Genotype Classification: Applies convolutional neural networks to classify structural variation genotypes from sequencing-derived representations.
• Sequencing Data Image Transformation: Converts next-generation sequencing data into image-based representations to enable visual pattern recognition by CNN models.
• Convolutional Bootstrapping Strategy: Implements convolutional bootstrapping to improve robustness and reduce the impact of noisy labels during genotype prediction.

Scientific Applications:

• Structural Variation Genotyping: Identifies and classifies structural variation genotypes from next-generation sequencing datasets.
• Genomic Variation Analysis: Supports studies of genetic diversity, disease-associated variants, and evolutionary genomics.

Methodology:

Cnngeno transforms next-generation sequencing data into image-based representations, applies convolutional neural network models for genotype classification, and uses convolutional bootstrapping to mitigate the effects of noisy training labels.