Enhancing prediction accuracy of coronary artery disease through machine learning-driven genomic variant selection

Machine learning (ML) methods are increasingly becoming crucial in genome-wide association studies for identifying key genetic variants or SNPs that statistical methods might overlook. Statistical methods predominantly identify SNPs with notable effect sizes by conducting association tests on individual genetic variants, one at a time, to determine their relationship with the target phenotype. These genetic variants are then used to create polygenic risk scores (PRSs), estimating an individual's genetic risk for complex diseases like cancer or cardiovascular disorders. Unlike traditional methods, ML algorithms can identify groups of low-risk genetic variants that improve prediction accuracy when combined in a mathematical model. However, the application of ML strategies requires addressing the feature selection challenge to prevent overfitting. Moreover, ensuring the ML model depends on a concise set of genomic variants enhances its clinical applicability, where testing is feasible for only a limited number of SNPs. In this study, we introduce a robust pipeline that applies ML algorithms in combination with feature selection (ML-FS algorithms), aimed at identifying the most significant genomic variants associated with the coronary artery disease (CAD) phenotype. The proposed computational approach was tested on individuals from the UK Biobank, differentiating between CAD and non-CAD individuals within this extensive cohort, and benchmarked against standard PRS-based methodologies like LDpred2 and Lassosum. Our strategy incorporates cross-validation to ensure a more robust evaluation of genomic variant-based prediction models. This method is commonly applied in machine learning strategies but has often been neglected in previous studies assessing the predictive performance of polygenic risk scores. Our results demonstrate that the ML-FS algorithm can identify panels with as few as 50 genetic markers that can achieve approximately 80% accuracy when used in combination with known risk factors. The modest increase in accuracy over PRS performances is noteworthy, especially considering that PRS models incorporate a substantially larger number of genetic variants. This extensive variant selection can pose practical challenges in clinical settings. Additionally, the proposed approach revealed novel CAD-genetic variant associations.

Network approaches for modeling the effect of drugs and diseases

The network approach is quickly becoming a fundamental building block of computational methods aiming at elucidating the mechanism of action (MoA) and therapeutic effect of drugs. By modeling the effect of drugs and diseases on different biological networks, it is possible to better explain the interplay between disease perturbations and drug targets as well as how drug compounds induce favorable biological responses and/or adverse effects. Omics technologies have been extensively used to generate the data needed to study the mechanisms of action of drugs and diseases. These data are often exploited to define condition-specific networks and to study whether drugs can reverse disease perturbations. In this review, we describe network data mining algorithms that are commonly used to study drug’s MoA and to improve our understanding of the basis of chronic diseases. These methods can support fundamental stages of the drug development process, including the identification of putative drug targets, the in silico screening of drug compounds and drug combinations for the treatment of diseases. We also discuss recent studies using biological and omics-driven networks to search for possible repurposed FDA-approved drug treatments for SARS-CoV-2 infections (COVID-19).

A toxicogenomic data space for system-level understanding and prediction of EDC-induced toxicity

Endocrine disrupting compounds (EDCs) are a persistent threat to humans and wildlife due to their ability to interfere with endocrine signaling pathways. Inspired by previous work to improve chemical hazard identification through the use of toxicogenomics data, we developed a genomic-oriented data space for profiling the molecular activity of EDCs in an in silico manner, and for creating predictive models that identify and prioritize EDCs. Predictive models of EDCs, derived from gene expression data from rats (in vivo and in vitro primary hepatocytes) and humans (in vitro primary hepatocytes and HepG2), achieve testing accuracy greater than 90%. Negative test sets indicate that known safer chemicals are not predicted as EDCs. The rat in vivo-based classifiers achieve accuracy greater than 75% when tested for invitro to in vivoextrapolation. This study reveals key metabolic pathways and genes affected by EDCs together with a set of predictive models that utilize these pathways to prioritize EDCs in dose/time dependent manner and to predict EDCevokedmetabolic diseases.

Feature set optimization in biomarker discovery from genome-scale data

MOTIVATION

Omics technologies have the potential to facilitate the discovery of new biomarkers. However, only few omics-derived biomarkers have been successfully translated into clinical applications to date. Feature selection is a crucial step in this process that identifies small sets of features with high predictive power. Models consisting of a limited number of features are not only more robust in analytical terms, but also ensure cost effectiveness and clinical translatability of new biomarker panels. Here we introduce GARBO, a novel multi-island adaptive genetic algorithm to simultaneously optimize accuracy and set size in omics-driven biomarker discovery problems.

RESULTS

Compared to existing methods, GARBO enables the identification of biomarker sets that best optimize the trade-off between classification accuracy and number of biomarkers. We tested GARBO and six alternative selection methods with two high relevant topics in precision medicine: cancer patient stratification and drug sensitivity prediction. We found multivariate biomarker models from different omics data types such as mRNA, miRNA, copy number variation, mutation and DNA methylation. The top performing models were evaluated by using two different strategies: the Pareto-based selection, and the weighted sum between accuracy and set size (w = 0.5). Pareto-based preferences show the ability of the proposed algorithm to search minimal subsets of relevant features that can be used to model accurate random forest-based classification systems. Moreover, GARBO systematically identified, on larger omics data types, such as gene expression and DNA methylation, biomarker panels exhibiting higher classification accuracy or employing a number of features much lower than those discovered with other methods. These results were confirmed on independent datasets.

AVAILABILITY AND IMPLEMENTATION

github.com/Greco-Lab/GARBO.

CONTACT

dario.greco@tuni.fi.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Antiproliferative and survival properties of PMA in MCF-7 breast cancer cell

Although PKCs are assumed to be the main targets of phorbol esters (PMA), additional PMA effectors, such as chimaerins (a family of RacGTPase activating proteins) and RasGRP (exchange factor for Ras/Rap1), can counteract or strengthen the PKC pathways. In this study, we evaluated the proliferative behavior of PMA-treated MCF-7 breast cancer cell and found that: PMA induced growth arrest and inhibited cell death; PMA activated ERKs, which, in turn, induced p21; and inhibitors of ERK (PD98059) and PKC (GF109203X) prevented p21 induction and abolished the PMA survival effect. We conclude that PMA inhibits MCF-7 cell growth and simultaneously stimulates cell survival; both responses are linked to ERK-dependent and p53-independent p21 induction.

Rottlerin inhibits the nuclear factor kappaB/cyclin-D1 cascade in MCF-7 breast cancer cells

In the course of a project aimed to clarify the molecular mechanisms by which phorbol 12-myristate 13-acetate (PMA)-activated forms of protein kinase C (PKC) promote growth arrest in an MCF-7 cell line, we found that the PKCdelta inhibitor Rottlerin was able by itself to block cell proliferation. In the current study, we investigated further the antiproliferative response to Rottlerin. Western blotting analysis of cytoplasmic/nuclear extracts showed that the drug did not prevent either extracellular signal-regulated kinase (ERK) activation by PMA or Akt phosphorylation, but did interfere with the NFkappaB activation process (both basal and PMA-stimulated), by lowering the levels of phospho-IkappaBalpha and preventing p65 nuclear migration. The growth arrest evoked by Rottlerin was not mediated by cell-cycle inhibitors p21 and p27 but was accompanied by a dramatic fall in the cyclin-D1 protein, the levels of which were not altered by the pan-PKC inhibitor GF 109203X, thus excluding a PKC-mediated mechanism in the Rottlerin effect. The parallel drop in cyclin-D1 mRNA suggested a down-regulation of the gene caused by the inhibition of nuclear factor-kappa B (NFkappaB), which occurs via a PKC-, Akt-, ERK- and mitochondrial uncoupling-independent mechanism. We provide preliminary evidence that the interference on the NFkappaB activation process likely occurs at the level of calcium/calmodulin-dependent protein kinase II (CaMKII), a known Rottlerin target. Indeed the drug prevented calcium-induced CaMKII autophosphorylation which, in turn, led to decreased NFkappaB activation.

Functional interactions of protein kinase A and C in signalling networks: a recapitulation

On the basis of evidence collected from the literature, we propose a general model by which protein kinase (PK) A and the different PKC isoforms can inversely affect cell growth. Molecular switches, which are able to direct the signal towards antiproliferative or mitogenic pathways, are the different isoforms of Raf and PKC. Conflicting data are also reported and discussed in an attempt to reconcile them.

The dual action of ozone on the skin

The aim of this brief review is to summarize the recent literature on the effect of ozone (O3) on cutaneous tissues. Recently it has been reported that a chronic contact with O3 can be deleterious for the skin. Our group and others have shown a progressive depletion of antioxidant content in the stratum corneum and this can then lead to a cascade of effects resulting in an active cellular response in the deeper layers of the skin. Using an in vivo model we have shown an increase of proliferative, adaptive and proinflammatory cutaneous tissue responses. On the other hand the well known activity of O3 as a potent disinfectant and oxygen (O2) donor has been also studied for therapeutic use. Two approaches have been described. The first consists of a quasi-total body exposure in a thermostatically controlled cabin. This treatment has proved to be useful in patients with chronic limb ischaemia. The second approach is based on the topical application of ozonated olive oil in several kinds of skin infection (from soreness to diabetic ulcers, burns, traumatic and surgical wounds, abscesses and skin reactions after radiotherapy). We and other authors have observed a striking cleansing effect with improved oxygenation and enhanced healing of these conditions. It is now clear that, on the skin, O3, like other drugs, poisons and radiation, can display either a damaging effect from a long exposure or a beneficial effect after a brief exposure to O2 and O3 or to the application of ozonated oil to chronic wounds.

The complexity of parathyroid hormone-related protein signalling

Our understanding of the mode of action of parathyroid hormone-related protein (PTHrP) has changed profoundly during the last decade. Most PTHrP activities are mediated by membrane receptors through autocrine/paracrine pathways. However, both endogenous and exogenous PTHrP also appear to have intracrine effects through translocation into the nucleus. The present review proposes unconventional PTHrP signalling, based on novel clues. First, PTHrP binding to its membrane receptor triggers internalization of the whole complex, mediated by beta-arrestin. There is growing evidence that the receptor and arrestin are the effectors of biological responses, rather than the ligand (or in addition to the ligand). Second, the existence of putative PTHrP targets within the cytoplasm is beginning to be supported. Recent findings of interactions between a COOH-terminus of PTHrP and beta-arrestin and between the PTHrP receptor and 14-3-3 proteins represent the starting point for identification of intracellular partners of both the hormone and its receptor.

ERKs are the point of divergence of PKA and PKC activation by PTHrP in human skin fibroblasts

Parathyroid hormone-related peptide (PTHrP) receptors, coupled to trimeric G proteins, operate in most target cells through at least three different transduction routes: Galpha s-mediated stimulation of adenylylcyclase (AC), Galpha q-mediated activation of phospholipase Cbeta (PLC) and mitogen-activated protein kinase (MAPK) activation. In this study we investigated the relative role of different pathways in human skin fibroblast proliferation. Using chemical inhibitors and activators of signal transduction, we demonstrated that: (i) AC/cAMP and PLC/1,4,5 inositol triphosphate/diacylglycerol second-messenger systems are simultaneously activated following PTHrP binding to its receptors; (ii) the mitogenic response to PTHrP derives from a balance between two counteracting pathways--an activating route mediated by protein kinase C (PKC) and an inhibitory route mediated by protein kinase A (PKA); (iii) PTHrP mitogenic effects are largely dependent on MAPKs, whose activity can be modulated by both PKA and PKC. Our results indicate that MAPKs are common targets of both transduction routes and, at the same time, their point of divergence in mediating PTHrP dual and opposite mitogenic effects.

Distribution of type-I interferon-receptors in human first trimester and term placental tissues and on isolated trophoblast cells

PROBLEM

Type-I interferon (IFN) is the protein recognizing pregnancy in ruminants. Although IFN is secreted in early pregnancy, its role is not still clear in other species. Like other cytokines, IFN exerts its biological functions through specific membrane receptors. We have investigated the potential action of IFN in human pregnancy by studying the distribution of the receptors in the human placenta.

METHOD

Reactivity to monoclonal antibodies (mAbs) to the type-I IFN-receptor (R) was analyzed by immunohistochemistry in human placental tissues and in cytospins of first trimester trophoblast cells.

RESULTS

Type-I IFN-R immunoreactivity was observed mostly in first trimester villous cytotrophoblasts and in the cytotrophoblast cell columns. Trophoblast in the decidua, the epithelium of the uterine glands, and most of the isolated trophoblast cells were also immunoreactive.

CONCLUSION

The expression of type-I IFN-R in the highly proliferating and migrating trophoblast suggests that this cytokine has a role in trophoblast growth and invasion.