PoCos: Population Covering Locus Sets for Risk Assessment in Complex Diseases

Abstract

Susceptibility loci identified by GWAS generally account for a limited fraction of heritability. Predictive models based on identified loci also have modest success in risk assessment and therefore are of limited practical use. Many methods have been developed to overcome these limitations by incorporating prior biological knowledge. However, most of the information utilized by these methods is at the level of genes, limiting analyses to variants that are in or proximate to coding regions. We propose a new method that integrates protein protein interaction (PPI) as well as expression quantitative trait loci (eQTL) data to identify sets of functionally related loci that are collectively associated with a trait of interest. We call such sets of loci “population covering locus sets” (PoCos). The contributions of the proposed approach are three-fold: 1) We consider all possible genotype models for each locus, thereby enabling identification of combinatorial relationships between multiple loci. 2) We develop a framework for the integration of PPI and eQTL into a heterogenous network model, enabling efficient identification of functionally related variants that are associated with the disease. 3) We develop a novel method to integrate the genotypes of multiple loci in a PoCo into a representative genotype to be used in risk assessment. We test the proposed framework in the context of risk assessment for seven complex diseases, type 1 diabetes (T1D), type 2 diabetes (T2D), psoriasis (PS), bipolar disorder (BD), coronary artery disease (CAD), hypertension (HT), and multiple sclerosis (MS). Our results show that the proposed method significantly outperforms individual variant based risk assessment models as well as the state-of-the-art polygenic score. We also show that incorporation of eQTL data improves the performance of identified POCOs in risk assessment. We also assess the biological relevance of PoCos for three diseases that have similar biological mechanisms and identify novel candidate genes. The resulting software is publicly available at http://compbio.case.edu/pocos/.

Prediction of Kinase-Substrate Associations Using The Functional Landscape of Kinases and Phosphorylation Sites

“… 18 While co-occurrence captures the relationship between pairs of sites that tend to appear in similar contexts at a broader scale, Co-P captures finer-scale correlations between the dynamic ranges of the phosphorylation levels of site pairs. To incorporate Co-P in the site association network, we use data from 9 mass spectrometry-based phosphoproteomic studied that represent a broad range of biological states and provide sufficient number of samples to enable reliable assessment of Co-P. 9 These datasets include data from three breast cancer studies, 19 – 21 two ovarian cancer studies, 20 , 22 one colorectal cancer, 23 one lung cancer, 24 one Alzheimer’s disease 25 and one retinal pigmented eputhelium data. 26 …”

“…The earlier KSA prediction methods focus mainly on sequence motifs recognized by the active sites of kinases.. 2 – 4 Later methods integrate other contextual information such as protein structure and physical interactions to improve the accuracy of prediction methods. 5 – 8 Recently, we developed CophosK, 9 the first kinase-substrate prediction algorithm that utilizes large-scale mass spectrometry based phospho-proteomic data to incorporate contextual information. While these tools improve the kinase-substrate associations prediction, the knowledge about the substrates of kinases is still unequally distributed, where 87% of phosphosites are assigned to 20% of well-studied kinases.…”

Uncovering complementary sets of variants for predicting quantitative phenotypes

“…While there are a number of approaches to choose from for binary phenotypes (e.g., as in case/control studies) (Yu and Liu, 2004;Ding and Peng, 2005;Guestrin et al, 2005;Van Hulse et al, 2012;Ayati and Koyutürk, 2016;Urbanowicz et al, 2018), the options for the SNP selection problem in quantitative phenotypes are more limited. In the machine learning community, this problem essentially translates to a feature selection task for regression analysis and there are many established methods for this purpose.…”

Potpourri: An Epistasis Test Prioritization Algorithm via Diverse SNP Selection

“…That is, each case sample has to have at least one SNP in each POCO. Epistasis tests then are performed across POCOs with the hope that independent coverage of the cases will lead different POCOs to include complementary and epistatic SNPs (18,19). Finally, Cowman and Koyuturk, introduced the LINDEN algorithm (20).…”