Back

Distinct pancreatic and neuronal Lung Carcinoid molecular subtypes revealed by integrative omic analysis

Domingo-Sabugo, C.; Willis-Owen, S. A. G.; Mandal, A.; Nastase, A.; Dwyer, S.; Brambilla, C.; Hector Galvez, J.; Zhuang, Q.; Popat, S.; Eveleigh, R.; Munter, M.; Lim, E.; G. Nicholson, A.; Lathrop, M.; Cookson, W. O. C.; Moffatt, M. F.

2021-07-30 genetic and genomic medicine
10.1101/2021.07.27.21260865 medRxiv
Show abstract

Lung Carcinoids (L-CDs) are uncommon low-grade neuroendocrine tumours that are only recently becoming characterised at the molecular level. Notably data on the molecular events that precipitate altered gene expression programmes are very limited. Here we have identified two discrete L-CD subtypes from transcriptomic and whole-genome DNA methylation data, and comprehensively defined their molecular profiles using Whole-Exome Sequencing (WES) and Single Nucleotide Polymorphism (SNP) genotyping. Subtype (Group) 1 features upregulation of neuronal markers (L-CD-NeU) and is characterised by focal spindle cell morphology, peripheral location (71%), high mutational load (P=3.4x10-4), recurrent copy number alterations and is enriched for Atypical Lung Carcinoids. Group 2 (L-CD-PanC) are centrally located and feature upregulation of pancreatic and metabolic pathway genes concordant with promoter hypomethylation of beta cell and genes related to insulin secretion (P<1x10-6). L-CD-NeU tumours harbour mutations in chromatin remodelling and in SWI/SNF complex members, while L-CD-PanC tumours show aflatoxin mutational signatures and significant DNA methylation loss genome-wide, particularly enriched in repetitive elements (P<2.2 x 10-16). Our findings provide novel insights into the distinct mechanisms of epigenetic dysregulation in these lung malignancies, potentially opening new avenues for biomarker selection and treatment in L-CD patients.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.