Topological data analysis identifies molecular phenotypes of idiopathic pulmonary fibrosis

Abstract

BackgroundIdiopathic pulmonary fibrosis (IPF) is a debilitating, progressive disease with a median survival time of 3–5 years. Diagnosis remains challenging and disease progression varies greatly, suggesting the possibility of distinct subphenotypes.Methods and resultsWe analysed publicly available peripheral blood mononuclear cell expression datasets for 219 IPF, 411 asthma, 362 tuberculosis, 151 healthy, 92 HIV and 83 other disease samples, totalling 1318 patients. We integrated the datasets and split them into train (n=871) and test (n=477) cohorts to investigate the utility of a machine learning model (support vector machine) for predicting IPF. A panel of 44 genes predicted IPF in a background of healthy, tuberculosis, HIV and asthma with an area under the curve of 0.9464, corresponding to a sensitivity of 0.865 and a specificity of 0.89. We then applied topological data analysis to investigate the possibility of subphenotypes within IPF. We identified five molecular subphenotypes of IPF, one of which corresponded to a phenotype enriched for death/transplant. The subphenotypes were molecularly characterised using bioinformatic and pathway analysis tools identifying distinct subphenotype features including one which suggests an extrapulmonary or systemic fibrotic disease.ConclusionsIntegration of multiple datasets, from the same tissue, enabled the development of a model to accurately predict IPF using a panel of 44 genes. Furthermore, topological data analysis identified distinct subphenotypes of patients with IPF which were defined by differences in molecular pathobiology and clinical characteristics.

How to handle big data for disease stratification in respiratory medicine?

“…The findings reported by Shapanis et al 16 and Kraven et al 18 are broadly consistent. It is worth noting that both studies had a number of datasets in commonhighlighting how extremely similar data may be analysed in different ways for differing purposes; Shapanis et al 16 identify a 44 gene transcriptional signature which has diagnostic utility, while Kraven et al 18 generated a 13 gene signature with prognostic potential. Employing different methodologies to identify subgroups of IPF patients may, therefore, explain the different numbers of subgroups and specific differentially expressed genes identified.…”

“…15 The study by Shapanis et al in Thorax used the TDA Mapper algorithm for stratification of idiopathic pulmonary fibrosis (IPF) patients. 16 IPF disease progression is known to be heterogeneous,…”

“…In contrast, a recent study by Kraven et al 18 using PCA, identified three clusters (subphenotypes) of IPF patients based on six independent whole blood gene expression datasets. Both studies by Shapanis et al 16 and Kraven et al 18 mapped clinical features on to the clusters and observed phenotypes with significantly different diffusing capacity of the lung for carbon monoxide. The TDA-constructed network identified a cluster (subphenotype 5) characterised by higher likelihood of death/transplant compared with the others.…”

“…This study was more limited in the availability of longitudinal clinical parameters with which to map against the endotypes, in contrast to. 16 In addition, the validation of the endotypes was carried out using blood transcriptomic datasets since no other BAL transcriptomic datasets were available; two of those datasets were also used by Shapanis et al 16 . It is important to note that these tissue compartments (BAL and blood) clearly have distinct cellular compositions, which will have significant influence-ideally the BAL endotypes would benefit from validation in further BAL gene expression datasets.…”

Persistent Homology Identifies Pathways Associated with Hepatocellular Carcinoma from Peripheral Blood Samples

“…In the realm of infectious diseases, TDA has been applied to whole-blood RNA sequencing (RNA-seq) gene expression data [18] and to measurements of inflammatory biomarkers in blood plasma [19] from coronavirus disease 2019 (COVID-19) patients. A TDA-based approach also demonstrated its capability to distinguish idiopathic pulmonary fibrosis from multiple other diseases based on peripheral blood mononuclear cell (PBMC) gene expression data [20].…”