A Topic Modeling Analysis of TCGA Breast and Lung Cancer Transcriptomic Data

Abstract

Topic modeling is a widely used technique to extract relevant information from large arrays of data. The problem of finding a topic structure in a dataset was recently recognized to be analogous to the community detection problem in network theory. Leveraging on this analogy, a new class of topic modeling strategies has been introduced to overcome some of the limitations of classical methods. This paper applies these recent ideas to TCGA transcriptomic data on breast and lung cancer. The established cancer subtype organization is well reconstructed in the inferred latent topic structure. Moreover, we identify specific topics that are enriched in genes known to play a role in the corresponding disease and are strongly related to the survival probability of patients. Finally, we show that a simple neural network classifier operating in the low dimensional topic space is able to predict with high accuracy the cancer subtype of a test expression sample.

Multi-omics Topic Modeling for Breast Cancer Classification

“…While the above results were similar to the ones already discussed in [3] , the novelty of the present analysis is that we can perform a similar study also on the miRNA side. As we shall see this allows to have a new independent insight on the problem.…”

“…Then, by subtracting to P(topic|subtype) the mean value over the whole dataset we find a new set of quantities, that we define as "centered" distributions (we already used them in [3] and it has the same meaning of the normalised value of the mixture proportion τ in [23])…”

Topic Evolution Analysis for Omics Data Integration in Cancers

“…LDA was first introduced by Blei's study in 2003 (Hofmann, 1999 ; Blei et al, 2003 ). Scholars in cell and developmental biology have applied the LDA model to identify scientific research topics (Li et al, 2015 ; Valle et al, 2020 ). Besides, perplexity is considered as a standard tool to evaluate the effectiveness of various natural language processing models (Rosen-Zvi et al, 2010 ).…”

Artificial Intelligence in Bulk and Single-Cell RNA-Sequencing Data to Foster Precision Oncology

“…In cancer transcriptomics, ML and DL models have been applied to classify different cancer subtypes and cell populations [ 17 , 18 , 19 , 20 ], characterize tumor immune microenvironment [ 21 , 22 , 23 , 24 , 25 ], discover new prognostic biomarkers [ 26 , 27 , 28 ], assess and predict disease recurrence and patient survival [ 29 , 30 , 31 , 32 ], identify new putative actionable vulnerabilities [ 33 , 34 ], and predict tumor antigen immunogenicity [ 35 ] ( Figure 2 ).…”

“…Developed for natural language processing, this probabilistic clustering algorithm aims at discovering the hidden “topics” that reflect the biological heterogeneity and enhancing its comprehensive interpretation [ 109 ]. Applied to breast and lung cancer RNA-seq datasets, topic modeling outperformed standard clustering algorithms in identifying subtype-specific molecular features and their corresponding clinical outcomes [ 20 ].…”