Are There Limits in Explainability of Prognostic Biomarkers? Scrutinizing Biological Utility of Established Signatures.

Prognostic biomarkers can have an important role in the clinical practice because they allow stratification of patients in terms of predicting the outcome of a disorder. Obstacles for developing such markers include lack of robustness when using different data sets and limited concordance among similar signatures. In this paper, we highlight a new problem that relates to the biological meaning of already established prognostic gene expression signatures. Specifically, it is commonly assumed that prognostic markers provide sensible biological information and molecular explanations about the underlying disorder. However, recent studies on prognostic biomarkers investigating 80 established signatures of breast and prostate cancer demonstrated that this is not the case. We will show that this surprising result is related to the distinction between causal models and predictive models and the obfuscating usage of these models in the biomedical literature. Furthermore, we suggest a falsification procedure for studies aiming to establish a prognostic signature to safeguard against false expectations with respect to biological utility.

Limitations of Explainability for Established Prognostic Biomarkers of Prostate Cancer.

High-throughput technologies do not only provide novel means for basic biological research but also for clinical applications in hospitals. For instance, the usage of gene expression profiles as prognostic biomarkers for predicting cancer progression has found widespread interest. Aside from predicting the progression of patients, it is generally believed that such prognostic biomarkers also provide valuable information about disease mechanisms and the underlying molecular processes that are causal for a disorder. However, the latter assumption has been challenged. In this paper, we study this problem for prostate cancer. Specifically, we investigate a large number of previously published prognostic signatures of prostate cancer based on gene expression profiles and show that none of these can provide unique information about the underlying disease etiology of prostate cancer. Hence, our analysis reveals that none of the studied signatures has a sensible biological meaning. Overall, this shows that all studied prognostic signatures are merely black-box models allowing sensible predictions of prostate cancer outcome but are not capable of providing causal explanations to enhance the understanding of prostate cancer.

Prognostic gene expression signatures of breast cancer are lacking a sensible biological meaning.

The identification of prognostic biomarkers for predicting cancer progression is an important problem for two reasons. First, such biomarkers find practical application in a clinical context for the treatment of patients. Second, interrogation of the biomarkers themselves is assumed to lead to novel insights of disease mechanisms and the underlying molecular processes that cause the pathological behavior. For breast cancer, many signatures based on gene expression values have been reported to be associated with overall survival. Consequently, such signatures have been used for suggesting biological explanations of breast cancer and drug mechanisms. In this paper, we demonstrate for a large number of breast cancer signatures that such an implication is not justified. Our approach eliminates systematically all traces of biological meaning of signature genes and shows that among the remaining genes, surrogate gene sets can be formed with indistinguishable prognostic prediction capabilities and opposite biological meaning. Hence, our results demonstrate that none of the studied signatures has a sensible biological interpretation or meaning with respect to disease etiology. Overall, this shows that prognostic signatures are black-box models with sensible predictions of breast cancer outcome but no value for revealing causal connections. Furthermore, we show that the number of such surrogate gene sets is not small but very large.

Graph-based exploitation of gene ontology using GOxploreR for scrutinizing biological significance.

Gene ontology (GO) is an eminent knowledge base frequently used for providing biological interpretations for the analysis of genes or gene sets from biological, medical and clinical problems. Unfortunately, the interpretation of such results is challenging due to the large number of GO terms, their hierarchical and connected organization as directed acyclic graphs (DAGs) and the lack of tools allowing to exploit this structural information explicitly. For this reason, we developed the R package GOxploreR. The main features of GOxploreR are (I) easy and direct access to structural features of GO, (II) structure-based ranking of GO-terms, (III) mapping to reduced GO-DAGs including visualization capabilities and (IV) prioritizing of GO-terms. The underlying idea of GOxploreR is to exploit a graph-theoretical perspective of GO as manifested by its DAG-structure and the containing hierarchy levels for cumulating semantic information. That means all these features enhance the utilization of structural information of GO and complement existing analysis tools. Overall, GOxploreR provides exploratory as well as confirmatory tools for complementing any kind of analysis resulting in a list of GO-terms, e.g., from differentially expressed genes or gene sets, GWAS or biomarkers. Our R package GOxploreR is freely available from CRAN.