ALPACA: a visual data mining system for subcellular location-specific knowledge mining from multi-omics data in cancer

Compartments such as subcellular organelles provide adequate environments and interacting partners to the constituent proteins so that their dedicated functions can be carried out properly. While mRNA or protein abundance data is commonly used in computational analysis, data mining from a single abundance data often do not provide sufficient information towards diseases like cancer that introduce various perturbations to biological functions. In addition, transcriptomic and proteomic data do not agree to a large degree. Such challenges make it imperative to comparatively analyze data of multiple omics in terms of subcellular locations to better understand the functional dynamics of cancer. In this paper, A Location-wise Proteome/transcriptome Abundance Comparative Analyzer (ALPACA) is introduced, a web-based visual data mining system for subcellular location-specific knowledge mining from multi-omics data.

ALPACA calculates location enrichment distributions from cancer abundance data and visually presents these in a pathway-specific manner, helping researchers observe biological function discrepancies. The effectiveness of ALPACA was demonstrated by reproducing known cancer-related genes and pathways across various types of cancers. For example, ALPACA was able to identify overexpression of TP53 and underexpression of DIABLO/Smac in basal-like breast cancer, as well as high discrepancy in Fc gamma R-mediated phagocytosis pathway among gynaecological cancers with an implicated gene GSN.

ALPACA compartmentalizes the whole transcriptome and proteome abundance data subcellular location-wise and visually presents the differences of selected cancers using subcellular locations and biological pathways. The filtering and sorting functions of ALPACA add biological pathway context to the data analysis to help identify potential biomarkers. ALPACA is available at https://github.com/mwpak/ALPACA .