Epidemiological topology data analysis links severe COVID-19 to RAAS and hyperlipidemia associated metabolic syndrome conditions

Abstract

Motivation: The emergence of COVID-19 (C19) created incredible worldwide challenges but offers unique opportunities to understand the physiology of its risk factors and their interactions with complex disease conditions, such as metabolic syndrome. To address the challenges of discovering clinically relevant interactions, we employed a unique approach for epidemiological analysis powered by redescription-based topological data analysis (RTDA).

Results: Here, RTDA was applied to Explorys data to discover associations among severe C19 and metabolic syndrome. This approach was able to further explore the probative value of drug prescriptions to capture the involvement of RAAS and hypertension with C19, as well as modification of risk factor impact by hyperlipidemia (HL) on severe C19. RTDA found higher-order relationships between RAAS pathway and severe C19 along with demographic variables of age, gender, and comorbidities such as obesity, statin prescriptions, HL, chronic kidney failure, and disproportionately affecting Black individuals. RTDA combined with CuNA (cumulant-based network analysis) yielded a higher-order interaction network derived from cumulants that furthered supported the central role that RAAS plays. TDA techniques can provide a novel outlook beyond typical logistic regressions in epidemiology. From an observational cohort of electronic medical records, it can find out how RAAS drugs interact with comorbidities, such as hypertension and HL, of patients with severe bouts of C19. Where single variable association tests with outcome can struggle, TDA's higher-order interaction network between different variables enables the discovery of the comorbidities of a disease such as C19 work in concert.

Availability and implementation: Code for performing TDA/RTDA is available in https://github.com/IBM/Matilda and code for CuNA can be found in https://github.com/BiomedSciAI/Geno4SD/.

Supplementary information: Supplementary data are available at Bioinformatics online.

Figure 1.

Two sample cohorts were drawn from the Explorys system: a random sample and all C19 patients plus an equal sized randomly selected non-C19 control group. The total counts indicate the number after dropping missing entries. Abbreviations include Alzheimer’s Disease (AD), hypertension (HT), chronic kidney disease (CKD), hyperlipidemia (HL), cardiovascular disease (CVD), type II diabetes (T2D), pulmonary edema (PE), beta blocker (BB), calcium channel blocker (CCB), African Americans (AfAm).

Figure 2.

Stratified logistic regressions predicting (a) Black samples’ affinity drawn from the randomized cohort, (b) C19 infection from the C19 stratified cohort, (c) CVD in the randomized cohort, (d) HT in the randomized cohort, (e) Severe C19 in the Black C19 only cohort, (f) C19 from the C19 cohort, (g) C19 from the C19 cohort, using pooled RAAS drugs as a covariate, (h) C19 from the C19 cohort absent African Americans. Log odds ratios are plotted with 95% confidence intervals and colored red if P-value $<.05$, otherwise gray.

Figure 3.

(a) Jaccard distances between joint predicates. Severe C19 cluster is indicated with a gray bar (top). The primary homologous subclusters are indicated in blue, green, red, and orange. (b) Barcode plots for the Jaccard distances shown in panel (a). Abscissa is Jaccard threshold. Ordinate is homology group index ranked by Jaccard filtration threshold. The groups H0, H1, and H2 refer to dimensions 0, 1, and 2.

Figure 4.

Representative cycles corresponding to side-bar color codes in Fig. 3a selected to highlight inclusions involving severe C19 and three HT drugs (ACE inhibitors, beta blockers, and ARBs), as well as African Americans. (a) ACE inhibitor cycle, (b) CCB/CKD/age cycle, (c) CCB/obesity/HL cycle, and (d) African American cycle.

Figure 5.

Stratified logistic regression predicting severe C19 computed on C19 only patients from the C19 cohort with (a) and without (b) hyperlipidemia. Log odds ratios are plotted with 95% confidence intervals and colored red if P-value $<.05$, otherwise gray.

Figure 6.

Network representing significant, higher-order interactions between different features with nodes colored by their relative rank (gradient of brown to green corresponds to higher to lower rank) and edges colored by their respective pairwise odds ratios (gradient of light to dark corresponds to low to high OR) and edge width indicates the strength of the connection between them. The communities obtained from this network are marked by dashed lines.