|
Abstract:
|
Lung cancers cause 1,5 million casualties per year worldwide. Despite their heterogeneity, lung cancers are classified into two classes as small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), which include lung adenocarcinoma (LUAD) and lung squamous cell cancer (LUSC). SCLC is the most aggressive type of lung cancer reaching average survival rates of 5% after 5 years from the time of diagnosis. The lack of knowledge about the underlying tumorigenic mechanisms and the lack of effective treatments make the situation more dramatic for SCLC tumours. Our main goal is to obtain a specific splicing signature for SCLC that may provide novel molecular targets for prognosis and therapy. For this purpose, we used the iso-kTSP algorithm, a recently developed computational method able to identify transcript isoform changes across different samples. This comparison-based approach allows to classify RNA-seq data from different tumour or normal samples by establishing a decision rule based on the relative ordering in a ranking of isoform expression values. We applied this method to samples from different lung cancers: SCLC, LUAD and LUSC; and also to normal lung samples. Our results revealed a set of distinct alternative splicing patterns in SCLC with potential functional relevance. This work shows that identification of alterations in alternative splicing can shed light on the study of new molecular mechanisms to develop prognostic and therapeutic targets of SCLC tumours. |
