Genetic Determinants of Somatic Selection of Mutational Processes in 3,566 Human Cancers.

The somatic landscape of the cancer genome results from different mutational processes represented by distinct "mutational signatures." Although several mutagenic mechanisms are known to cause specific mutational signatures in cell lines, the variation of somatic mutational activities in patients, which is mostly attributed to somatic selection, is still poorly explained. Here, we introduce a quantitative trait, mutational propensity (MP), and describe an integrated method to infer genetic determinants of variations in the mutational processes in 3,566 cancers with specific underlying mechanisms. As a result, we report 2,314 candidate determinants with both significant germline and somatic effects on somatic selection of mutational processes, of which, 485 act via cancer gene expression and 1,427 act through the tumor-immune microenvironment. These data demonstrate that the genetic determinants of MPs provide complementary information to known cancer driver genes, clonal evolution, and clinical biomarkers. SIGNIFICANCE: The genetic determinants of the somatic mutational processes in cancer elucidate the biology underlying somatic selection and evolution of cancers and demonstrate complementary predictive power across cancer types.

Ratio of the interferon-γ signature to the immunosuppression signature predicts anti-PD-1 therapy response in melanoma.

Immune checkpoint inhibitor (ICI) treatments produce clinical benefit in many patients. However, better pretreatment predictive biomarkers for ICI are still needed to help match individual patients to the treatment most likely to be of benefit. Existing gene expression profiling (GEP)-based biomarkers for ICI are primarily focused on measuring a T cell-inflamed tumor microenvironment that contributes positively to the response to ICI. Here, we identified an immunosuppression signature (IMS) through analyzing RNA sequencing data from a combined discovery cohort (n = 120) consisting of three publicly available melanoma datasets. Using the ratio of an established IFN-γ signature and IMS led to consistently better prediction of the ICI therapy outcome compared to a collection of nine published GEP signatures from the literature on a newly generated internal validation cohort (n = 55) and three published datasets of metastatic melanoma treated with anti-PD-1 (n = 54) and anti-CTLA-4 (n = 42), as well as in patients with gastric cancer treated with anti-PD-1 (n = 45), demonstrating the potential utility of IMS as a predictive biomarker that complements existing GEP signatures for immunotherapy.

A machine learning framework to determine geolocations from metagenomic profiling.

BACKGROUND: Studies on metagenomic data of environmental microbial samples found that microbial communities seem to be geolocation-specific, and the microbiome abundance profile can be a differentiating feature to identify samples' geolocations. In this paper, we present a machine learning framework to determine the geolocations from metagenomics profiling of microbial samples. RESULTS: Our method was applied to the multi-source microbiome data from MetaSUB (The Metagenomics and Metadesign of Subways and Urban Biomes) International Consortium for the CAMDA 2019 Metagenomic Forensics Challenge (the Challenge). The goal of the Challenge is to predict the geographical origins of mystery samples by constructing microbiome fingerprints.First, we extracted features from metagenomic abundance profiles. We then randomly split the training data into training and validation sets and trained the prediction models on the training set. Prediction performance was evaluated on the validation set. By using logistic regression with L2 normalization, the prediction accuracy of the model reaches 86%, averaged over 100 random splits of training and validation datasets.The testing data consists of samples from cities that do not occur in the training data. To predict the "mystery" cities that are not sampled before for the testing data, we first defined biological coordinates for sampled cities based on the similarity of microbial samples from them. Then we performed affine transform on the map such that the distance between cities measures their biological difference rather than geographical distance. After that, we derived the probabilities of a given testing sample from unsampled cities based on its predicted probabilities on sampled cities using Kriging interpolation. Results show that this method can successfully assign high probabilities to the true cities-of-origin of testing samples. CONCLUSION: Our framework shows good performance in predicting the geographic origin of metagenomic samples for cities where training data are available. Furthermore, we demonstrate the potential of the proposed method to predict metagenomic samples' geolocations for samples from locations that are not in the training dataset.