Prior knowledge transfer across transcriptional data sets and technologies using compositional statistics yields new mislabelled ovarian cell line

Abstract

Here, we describe gene expression compositional assignment (GECA), a powerful, yet simple method based on compositional statistics that can validate the transfer of prior knowledge, such as gene lists, into independent data sets, platforms and technologies. Transcriptional profiling has been used to derive gene lists that stratify patients into prognostic molecular subgroups and assess biomarker performance in the pre-clinical setting. Archived public data sets are an invaluable resource for subsequent in silico validation, though their use can lead to data integration issues. We show that GECA can be used without the need for normalising expression levels between data sets and can outperform rank-based correlation methods. To validate GECA, we demonstrate its success in the cross-platform transfer of gene lists in different domains including: bladder cancer staging, tumour site of origin and mislabelled cell lines. We also show its effectiveness in transferring an epithelial ovarian cancer prognostic gene signature across technologies, from a microarray to a next-generation sequencing setting. In a final case study, we predict the tumour site of origin and histopathology of epithelial ovarian cancer cell lines. In particular, we identify and validate the commonly-used cell line OVCAR-5 as non-ovarian, being gastrointestinal in origin. GECA is available as an open-source R package.

Eliciting OTUD3/RIPK-Dependent Necroptosis to Prevent Epithelial Ovarian Cancer

“…Because in vivo use of TNFα would likely induce inflammation and other undesirable side effects [33], we chose to further examine if ESA reduces FTE or HGSOC viability. ESA treatment reduced cell viability of transformed FTE cells, as well as OVSAHO cells (HGSOC, TP53-mutant, BRCA2deletion, PGR positive), PEO1 cells (HGSOC, TP53-mutant, BRCA2-mutant, PGR negative), and OVCAR5 (likely a gastrointestinal carcinoma [34], TP53-mutant, BRCA-wildtype, PGR status unknown) ( Fig. 4A).…”

The Endometriotic Tumor Microenvironment in Ovarian Cancer

“…The number of cell lines derived from either endometrioid or clear cell ovarian cancers is more limited than high-grade serous cell lines. However, molecular profiling, including attention to gene mutations common in these endometriosis-associated ovarian cancers (i.e., ARID1A , PIK3CA , CTNNB1 , PTEN , and KRAS ) and mutations common in high-grade serous (i.e., TP53), have allowed better molecular and biological distinction [ 176 , 177 , 178 , 179 , 180 , 181 , 182 ]. Table 4 shows the common endometrioid and clear-cell ovarian cancer cell lines, including lines that were not derived from endometriosis-associated ovarian cancers, but which may molecularly represent non-high grade serous cell lines.…”

Compositional analysis gives insight into leukaemia cell lines expression profiles compared to those within patient sub‐groups

“…Using the power of gene expression profiling of a disease reference data we have robustly allocated a panel of commonly used leukaemia cell lines to disease sub‐groups. GECA, G ene E xpression C ompositional A ssignment (Blayney et al , ), is an in silico method using compositional statistics to transfer prior knowledge from a reference data set to an unclassified query set. The approach enables the comparison of similarities in gene expression profiles across independent datasets, platforms and technologies, thus removing cross platform normalisation.…”

“…The approach enables the comparison of similarities in gene expression profiles across independent datasets, platforms and technologies, thus removing cross platform normalisation. GECA had been applied to a library of epithelial ovarian cell lines with respect to a reference set of solid tumours (Blayney et al , ).…”

“…Cell lines by gene expression compositional aggregate assignment with heat maps showing percentage assignment of cell lines from each of the Query datasets 1, 2 and 3 compared to the reference data set (Haferlach et al , ) by GCEA (Blayney et al , ). AML , acute myeloid leukaemia; B‐ ALL , B cell acute lymphoblastic leukaemia; c‐ ALL , common type of acute lymphoblastic leukaemia; CLL , chronic lymphocytic leukaemia; CML , chronic myeloid leukaemia; MDS , myelodysplastic syndrome; Pre‐ BALL , precursor B‐ ALL ; Pro‐B‐ ALL , Early precursor B‐ ALL ; T‐ ALL , T cell acute lymphoblastic leukaemia.…”