Bayesian Framework for Detecting Gene Expression Outliers in Individual Samples

Abstract

PURPOSE Many antineoplastics are designed to target upregulated genes, but quantifying upregulation in a single patient sample requires an appropriate set of samples for comparison. In cancer, the most natural comparison set is unaffected samples from the matching tissue, but there are often too few available unaffected samples to overcome high intersample variance. Moreover, some cancer samples have misidentified tissues of origin or even composite-tissue phenotypes. Even if an appropriate comparison set can be identified, most differential expression tools are not designed to accommodate comparisons to a single patient sample. METHODS We propose a Bayesian statistical framework for gene expression outlier detection in single samples. Our method uses all available data to produce a consensus background distribution for each gene of interest without requiring the researcher to manually select a comparison set. The consensus distribution can then be used to quantify over- and underexpression. RESULTS We demonstrate this method on both simulated and real gene expression data. We show that it can robustly quantify overexpression, even when the set of comparison samples lacks ideally matched tissue samples. Furthermore, our results show that the method can identify appropriate comparison sets from samples of mixed lineage and rediscover numerous known gene-cancer expression patterns. CONCLUSION This exploratory method is suitable for identifying expression outliers from comparative RNA sequencing (RNA-seq) analysis for individual samples, and Treehouse, a pediatric precision medicine group that leverages RNA-seq to identify potential therapeutic leads for patients, plans to explore this method for processing its pediatric cohort.

Using machine learning approaches for multi-omics data analysis: A review

“…in the case of metabolomics, some values might be below the limit of detection and hence assigned null value (Antonelli et al, 2019)) than others. Therefore, imputation (Liew et al, 2011) and outlier detection (Vivian et al, 2020) should be considered for each omic separately, before planning their integration.…”

Applications of multi‐omics analysis in human diseases

“…In addition, some omics may produce null values due to being below the detection line of the instrument, such as metabolomics 93 . Therefore, before integrating multi‐omics data, the imputation 94 and outlier detection 95 for each omics data should be considered separately. Next, we briefly discuss some current common methods for multi‐omics integration (Figure 3).…”

Combining Molecular, Imaging, and Clinical Data Analysis for Predicting Cancer Prognosis

“…First, multi-omics data from high-throughput sources are generally heterogeneous and require pre-processing [50]. Among the most common pre-processing steps are normalization, scaling, imputation [51], and outlier detection techniques [52]. Imputation and outlier detection techniques need to be applied to each omics independently before proceeding to data analysis and integration [53].…”