MET Exon 14 Skipping: A Case Study for the Detection of Genetic Variants in Cancer Driver Genes by Deep Learning.
MET
deep learning
exon skipping
genetic variants
neural network
Journal
International journal of molecular sciences
ISSN: 1422-0067
Titre abrégé: Int J Mol Sci
Pays: Switzerland
ID NLM: 101092791
Informations de publication
Date de publication:
19 Apr 2021
19 Apr 2021
Historique:
received:
25
03
2021
revised:
13
04
2021
accepted:
17
04
2021
entrez:
30
4
2021
pubmed:
1
5
2021
medline:
25
5
2021
Statut:
epublish
Résumé
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. 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. 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. 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.
Sections du résumé
BACKGROUND
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
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
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
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.
Identifiants
pubmed: 33921709
pii: ijms22084217
doi: 10.3390/ijms22084217
pmc: PMC8072630
pii:
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Références
BMC Genomics. 2018 Dec 27;19(1):971
pubmed: 30591034
BMC Bioinformatics. 2019 Sep 13;20(1):469
pubmed: 31519146
BMC Bioinformatics. 2011 Aug 10;12:333
pubmed: 21831268
Lung Cancer. 2020 Feb;140:46-54
pubmed: 31862577
Curr Treat Options Oncol. 2020 Apr 18;21(4):33
pubmed: 32306194
BMC Syst Biol. 2019 Apr 5;13(Suppl 2):35
pubmed: 30953498
F1000Res. 2013 Sep 16;2:188
pubmed: 24555089
Cancer Discov. 2015 Aug;5(8):802-5
pubmed: 26243862
J Clin Oncol. 2016 Mar 1;34(7):721-30
pubmed: 26729443
Int J Cancer. 2018 Dec 1;143(11):2838-2848
pubmed: 30144023
Cancer Res. 2006 Jan 1;66(1):283-9
pubmed: 16397241
J Mol Biol. 1990 Oct 5;215(3):403-10
pubmed: 2231712
Nucleic Acids Res. 2020 Jan 8;48(D1):D896-D907
pubmed: 31642488
J Thorac Oncol. 2019 Apr;14(4):737-741
pubmed: 30639620
Nat Genet. 2008 Dec;40(12):1413-5
pubmed: 18978789
J Thorac Dis. 2020 May;12(5):2172-2178
pubmed: 32642122
NPJ Syst Biol Appl. 2021 Jan 5;7(1):1
pubmed: 33402683
Nat Rev Cancer. 2018 Nov;18(11):696-705
pubmed: 30293088
Biochem Biophys Res Commun. 2010 Apr 16;394(4):1042-6
pubmed: 20331976
Oncogene. 2017 Mar 2;36(9):1200-1210
pubmed: 27524418
N Engl J Med. 2020 Sep 3;383(10):931-943
pubmed: 32469185
Int J Mol Sci. 2020 Nov 28;21(23):
pubmed: 33260643
Biochim Biophys Acta. 2009 Jan;1792(1):14-26
pubmed: 18992329
Trends Genet. 2001 Feb;17(2):100-7
pubmed: 11173120
Cell. 2019 Jan 24;176(3):535-548.e24
pubmed: 30661751
Int J Radiat Oncol Biol Phys. 2012 Nov 15;84(4):e525-31
pubmed: 22836051
Bioinformatics. 2018 Dec 15;34(24):4180-4188
pubmed: 29931149
Gut. 2014 Apr;63(4):635-46
pubmed: 23704319
Nat Biotechnol. 2011 Jan;29(1):24-6
pubmed: 21221095
Clin Oncol (R Coll Radiol). 2016 Aug;28(8):e35-44
pubmed: 26880063
Cancer Res. 2007 Mar 1;67(5):2081-8
pubmed: 17332337
Comput Struct Biotechnol J. 2020 Dec 24;19:183-195
pubmed: 33425250
Front Bioeng Biotechnol. 2019 Nov 05;7:305
pubmed: 31750297
Oncotarget. 2015 Sep 29;6(29):28211-22
pubmed: 26375439
Nat Methods. 2018 Jul;15(7):539-542
pubmed: 29941873
Cancer Discov. 2015 Aug;5(8):850-9
pubmed: 25971938
BMC Cancer. 2019 Dec 5;19(1):1189
pubmed: 31805884
Int J Mol Sci. 2017 Dec 12;18(12):
pubmed: 29231888
J Thorac Oncol. 2016 Sep;11(9):1423-32
pubmed: 27223456