Related Experiment Video
Updated: Nov 7, 2025
Detecting Somatic Genetic Alterations in Tumor Specimens by Exon Capture and Massively Parallel Sequencing
Published on: October 18, 2013
MET Exon 14 Skipping: A Case Study for the Detection of Genetic Variants in Cancer Driver Genes by Deep Learning
Vladimir Nosi1, Alessandrì Luca1, Melissa Milan2
1Department of Molecular Biotechnology and Health Sciences, University of Torino, 10126 Torino, Italy.
Background:
Disruption of alternative splicing (AS) is frequently observed in cancer and might represent an important signature for tumor progression and therapy. Exon skipping (ES) represents one of the most frequent AS events, and in non-small cell lung cancer (NSCLC) MET exon 14 skipping was shown to be targetable.
Methods:
We constructed neural networks (NN/CNN) specifically designed to detect MET exon 14 skipping events using RNAseq data. Furthermore, for discovery purposes we also developed a sparsely connected autoencoder to identify uncharacterized MET isoforms.
Results:
The neural networks had a Met exon 14 skipping detection rate greater than 94% when tested on a manually curated set of 690 TCGA bronchus and lung samples. When globally applied to 2605 TCGA samples, we observed that the majority of false positives was characterized by a blurry coverage of exon 14, but interestingly they share a common coverage peak in the second intron and we speculate that this event could be the transcription signature of a LINE1 (Long Interspersed Nuclear Element 1)-MET (Mesenchymal Epithelial Transition receptor tyrosine kinase) fusion.
Conclusions:
Taken together, our results indicate that neural networks can be an effective tool to provide a quick classification of pathological transcription events, and sparsely connected autoencoders could represent the basis for the development of an effective discovery tool.
Insights
Neural networks effectively detect MET exon 14 skipping in non-small cell lung cancer (NSCLC), aiding in identifying targetable tumor progression signatures. Autoencoders also show promise for discovering novel MET isoforms.
Area of Science:
- Oncology
- Bioinformatics
- Genomics
Background:
- Alternative splicing (AS) disruptions are common in cancer, potentially indicating tumor progression and therapeutic response.
- Exon skipping (ES), a frequent AS event, includes MET exon 14 skipping in non-small cell lung cancer (NSCLC), which is a targetable alteration.
Purpose of the Study:
- To develop and validate neural networks (NN/CNN) for detecting MET exon 14 skipping events using RNA-seq data.
- To employ a sparsely connected autoencoder for discovering novel MET isoforms.
Main Methods:
- Construction of neural networks (NN/CNN) tailored for MET exon 14 skipping detection from RNA-seq data.
- Development of a sparsely connected autoencoder for identifying uncharacterized MET isoforms.
Main Results:
- Neural networks achieved over 94% detection rate for MET exon 14 skipping on a curated set of 690 TCGA lung cancer samples.
- Analysis of 2605 TCGA samples revealed common false positive patterns, suggesting a potential LINE1-MET fusion signature.
Conclusions:
- Neural networks offer an efficient method for classifying pathological transcription events.
- Sparsely connected autoencoders hold potential as a discovery tool for novel molecular signatures.
