Biology subjects

Halliwill, K.

Publications and source records attributed to Halliwill, K..

3 recordsLinked to original sources

bioRxiv2024

tauX: A Gene Expression Ratio Strategy to Improve Machine Learning Applications in Precision Medicine

Machine learning algorithms identify patterns that would otherwise be difficult to observe in high-dimensional molecular and clinical data. For this reason, machine learning has the potential to have a profound impact on clinical decision making and drug target discovery. Nevertheless, there remain considerable technical challenges in adapting these tools for clinical use. These challenges include clinical feature engineering, model selection, and defining optimal strategies for model training. For cancer care, RNA sequencing of patient tumor biopsies has already proven to be a powerful molecular assay to characterize tumor-intrinsic and -extrinsic phenotypes influencing therapeutic response. To improve the predictive performance of RNA-sequencing data, we developed the tauX machine learning framework to refine gene expression features and improve the performance of machine learning algorithms. The tauX framework uses aggregated ratios of positively and negatively associated predictive genes to simplify the prediction task. We showed a significant improvement in predictive performance using a large database of synthetic gene expression profiles. We also show how the tauX framework can be used to elucidate the mechanisms of response and resistance to checkpoint blockade therapy using data from the Stand Up to Cancer (SU2C) Lung Response Cohort and The Cancer Genome Atlas (TCGA). The tauX strategy achieved superior predictive performance compared to models built upon established feature engineering strategies or widely used cancer gene expression signatures. The tauX framework is available as a freely deployable docker container (https://hub.docker.com/r/pfeiljx/taux).

bioinformatics↗

bioRxiv2024

The interplay between dormant mutated cells and tumor promotion by chronic tissue damage in determining cancer risk.

While the causal role of mutagenic carcinogens in tumor development is well established, the relative contribution of environmental tumor promoting factors, wounding, and chronic inflammation is still unclear. Recent sequencing studies have suggested that most environmental carcinogens act as promoters rather than through mechanisms that involve direct induction of point mutations, but whether cancer risk factors such as obesity, chronic inflammation, wounding, or tumor promoters contribute directly or indirectly to mutation burden, or induce novel signatures, has not been investigated. Here, we present WGS analysis of over 100 mouse skin tumors to compare the effects of exposure to mutagens, the tumor promoter TPA, chronic wounding, obesity, or chemotherapy, on mutational burden and cancer risk. All tumors initiated by the carcinogen Dimethylbenzanthracene (DMBA) show a very strong A>T mutational signature (SBS.DMBA) attributable to a single exposure to this carcinogen. The number of SBS.DMBA mutations also showed a strong correlation with the "clock" signature SBS5, suggesting that one treatment with this mutagen can induce mutational signatures attributed to endogenous processes. No specific signatures could be attributed to obesity, high fat diet, wounding, or TPA. Cells carrying thousands of mutations persist over very long periods without inducing tumors or causing pathological changes but can give rise to tumors after short term exposure to TPA. Furthermore, normal cell turnover and proliferation during fetal and adult growth, is not sufficient for promotion, but tissue damage followed by regenerative proliferation seems to be required for tumor development. We conclude that tumor promoters, chronic inflammation, wounding, and obesity do not contribute significantly to tumor mutational burden, and that the rate-limiting determinant of tumor growth is exposure to a tumor promoter rather than the nature or number of genomic point mutations. These data are highly relevant to the recent demonstration of persistent oncogenic mutations in histologically normal human tissues during ageing.

cancer biology↗

bioRxiv2019

Conditional activation of immune-related signatures and prognostic significance: a pan-cancer analysis

BackgroundIt is becoming clear that tumor immune T cell infiltration and its functional orientation have substantial effect on cancer progression, influencing both response to therapy and prognosis. In this pan-cancer study, the previously described Immunologic Constant of Rejection (ICR) signature is used to define opposing immune phenotypes (i.e., immuneactive and immune-silent) across 31 different histologies. We systematically analyze the interconnections between the genetic programming of neoplasms and their immune orientation across different histologies, and the prognostic impact of such interplay. Moreover, we investigated the predictive value of ICR classification across various public datasets of immune checkpoint inhibition therapy. MethodsRNA-seq data of samples from a total of 9,282 patient tumor samples representing 31 cancer types were obtained from The Cancer Genome Atlas (TCGA). We classified each cancer type based on the expression of the ICR gene signature. Oncogenic pathway gene set enrichment and mutational status were analyzed in relation to ICR phenotypes. To explore whether tumorintrinsic attributes associate with the prognostic value of ICR across cancers, we compared mutational load, oncogenic alterations and expression of oncogenic pathways between cancer types using an integrative bioinformatic pipeline. ResultsOur analyses identified a distinct prognostic connotation of ICR depending on cancer histology. We identified several oncogenic pathways whose enrichment inversely correlated with ICR in multiple tumor types. We found several cancer specific pathways that were differentially enriched between tumors in which ICR had a prognostic impact versus the ones in which ICR did not bear any prognostic connotation such as proliferation and TGF-beta signaling. Importantly, this conditional impact of ICR was also validated in the context of immune checkpoint inhibition treatment. ConclusionsWe identified tumor-intrinsic attributes that correlate with immune phenotypes and potentially influence their development. In addition, a relationship was observed between the enrichment of oncogenic pathways and the prognostic significance of the ICR and its predictive value for patients treated with anti-CTLA4 immune checkpoint inhibition. Such information can be used to prioritize potential candidates for therapies aimed at converting immune-silent into immuneactive tumors and to refine stratification algorithms.

cancer biology↗