Tissue-aware data integration approach for the inference of pathway interactions in metazoan organisms

The majority of breast cancers are driven by the female hormone oestrogen via oestrogen receptor (ER) alpha. ER-positive patients are commonly treated with adjuvant endocrine therapy, however, resistance is a common occurrence and aside from ER-status, no unequivocal predictive biomarkers are currently in clinical use. In this study, we aimed to identify constitutional genetic variants influencing breast cancer survival among ER-positive patients and specifically, among endocrine-treated patients. We conducted a meta-analysis of three genome-wide association studies comprising in total 3,136 patients with ER-positive breast cancer of which 2,751 had received adjuvant endocrine therapy. We identified a novel locus (rs992531 at 8p21.2) associated with reduced survival among the patients with ER-positive breast cancer (P = 3.77 × 10−8). Another locus (rs7701292 at 5q21.3) was associated with reduced survival among the endocrine-treated patients (P = 2.13 × 10−8). Interaction analysis indicated that the survival association of rs7701292 is treatment-specific and independent of conventional prognostic markers. In silico functional studies suggest plausible biological mechanisms for the observed survival associations and a functional link between the putative target genes of the rs992531 and rs7701292 (RHOBTB2 and RAB9P1, respectively). We further explored the genetic interaction between rs992531 and rs7701292 and found a significant, treatment-specific interactive effect on survival among ER-positive, endocrine-treated patients (hazard ratio = 6.97; 95% confidence interval, 1.79–27.08, Pinteraction = 0.036). This is the first study to identify a genetic interaction that specifically predicts treatment outcome. These findings may provide predictive biomarkers based on germ line genotype informing more personalized treatment selection.

24Tuberculosis (TB) represents a global public health threat and is the leading cause of morbidity 25 and mortality worldwide. Effective control of TB is complicated with the emergence of multidrug 26resistance. Yet, there is still a fundamental gap in understanding the complex and dynamic 27 interactions between different Mycobacterium tuberculosis strains and the host. In this project, we 28 investigated the host immune response to different M. tuberculosis strains, including avirulent or 29 virulent and rifampin-resistant or isoniazid-resistant strains in THP-1 cells. We identified major 30 differences in the gene response profiles in response to infection with these strains. The expression 31 of IDO1 and IL-1b in the infected cells was stronger in all virulent M. tuberculosis strains. The 32 most striking result was the overexpression of many interferon-stimulated genes (ISGs) in cells 33 infected with the isoniazid-resistant strain, compared to the rifampin-drug resistant strain and the 34 drug-sensitive strain. A transcription regulation analysis of the differentially expressed genes in 35 infected THP-1 cells implicated two major transcription factors, NF-kB and STAT1. The 36 differentially expressed ISGs in response to the isoniazid-resistant M. tuberculosis strain were 37 associated with STAT1 signaling, while the expression of many cytokines, such IL-1b, was 38 associated with NF-kB signaling. Our data suggest that the isoniazid-resistant M. tuberculosis 39 strain preferentially activates STAT1 in response to cGAS-STING activation and induces a host 40 immune response signature that is characteristic of isoniazid resistance. This study has a potential 41 to provide important new insights into TB pathogenesis and to characterize host gene signatures 42 specifically involved in isoniazid-resistant TB. 43 44 45 groups (24 h post-infection; p<0.05). The 20 highly significant differentially expressed genes in 439 each plot are indicated in the insets. 440 441 442 443 H37Rv vs H37Rv-INH-R H37Rv-Rif-R vs H37Rv-INH-R MX1,

networks are largely used for modelling and analysing a wide range of biological data. As a consequence, many different research efforts have resulted in the introduction of a large number of algorithms for analysis and comparison of networks. Many of these algorithms can deal with networks with a single class of nodes and edges, also referred to as homogeneous networks. Recently, many different approaches tried to integrate into a single model the interplay of different molecules. A possible formalism to model such a scenario comes from node/edge coloured networks (also known as heterogeneous networks) implemented as node/ edge-coloured graphs. therefore, the need for the introduction of algorithms able to compare heterogeneous networks arises. We here focus on the local comparison of heterogeneous networks, and we formulate it as a network alignment problem. to the best of our knowledge, the local alignment of heterogeneous networks has not been explored in the past. We here propose L-HetnetAligner a novel algorithm that receives as input two heterogeneous networks (node-coloured graphs) and builds a local alignment of them. We also implemented and tested our algorithm. Our results confirm that our method builds high-quality alignments. The following website *contains Supplementary File 1 material and the code.Network Alignment (GNA) algorithms try to find a global mapping among all the nodes of the input networks, while Local Network Alignment (LNA) algorithms focus on mapping among (relatively) small single regions of input networks 12 . LNA has been defined in the past for homogeneous networks (LNA hom ), and it has been formalised in many papers, such as the first paper by Berg and Lassig 13 and the different formalisation proposed by Mina and Guzzi 14 . LNA algorithms try to find a mapping among (small) subregions of the input graphs 14 .Despite the existence of many algorithms for the local alignment of homogeneous networks 12 (see related work section for a detailed synopsis), they are not able to deal with heterogeneous networks since existing algorithms may process only homogeneous networks. Therefore they fail to discriminate among different node types. The alignment of heterogeneous networks is a relatively new field; Gu et al. 15 presented a novel GNA algorithm for heterogeneous networks, while to the best of our knowledge there are no available LNA algorithms designed for heterogeneous networks. Since the local alignment of networks reveals different knowledge compared to global alignment, there is a need for the introduction of novel LNA algorithms for heterogeneous networks.Here we propose L-HetNetAligner, a novel algorithm for local alignment of heterogeneous networks by proposing a two-step strategy as depicted in Fig. 2. Our algorithm takes as input two heterogeneous networks modelled as node-coloured graphs and a set of initial similarities among nodes of the networks, and it produces a set of graphs representing single local regions of the alignment. The algorithm merges two input graphs int...