Interpretable machine learning identifies paediatric Systemic Lupus Erythematosus subtypes based on gene expression data

Abstract

Transcriptomic analyses are commonly used to identify differentially expressed genes between patients and controls, or within individuals across disease courses. These methods, whilst effective, cannot encompass the combinatorial effects of genes driving disease. We applied rule-based machine learning (RBML) models and rule networks (RN) to an existing paediatric Systemic Lupus Erythematosus (SLE) blood expression dataset, with the goal of developing gene networks to separate low and high disease activity (DA1 and DA3). The resultant model had an 81% accuracy to distinguish between DA1 and DA3, with unsupervised hierarchical clustering revealing additional subgroups indicative of the immune axis involved or state of disease flare. These subgroups correlated with clinical variables, suggesting that the gene sets identified may further the understanding of gene networks that act in concert to drive disease progression. This included roles for genes (i) induced by interferons (IFI35 and OTOF), (ii) key to SLE cell types (KLRB1 encoding CD161), or (iii) with roles in autophagy and NF-κB pathway responses (CKAP4). As demonstrated here, RBML approaches have the potential to reveal novel gene patterns from within a heterogeneous disease, facilitating patient clinical and therapeutic stratification.

Recent advances in the use of machine learning and artificial intelligence to improve diagnosis, predict flares, and enrich clinical trials in lupus

“…Ultimately, they were able to classify patients with high and low disease activity using the expression of only 34 genes (accuracy = 81%). Examination of the genes that discriminated each subset provided useful information about underlying subset biology/pathogenesis, including the finding that the interferon response term was limited to low disease activity in this cohort [22 ▪ ]. Hierarchical clustering of the initial machine learning model output resulted in even further granularity about the disease activity subtypes, which were correlated with clinical data.…”

“…Yones et al [22 ▪ ] applied machine learning to blood gene expression data from pediatric patients with lupus with the aim of classifying patients into those with high and low disease activity. The authors repeatedly reduced the number of model input genes, by various feature selection techniques.…”

What is circulating factor disease and how is it currently explained?

“…These approaches allow us to take a data-driven, hypothesisfree approach to the analysis of huge biological datasets. This methodology is being utilised with great effect, particularly in oncology, with a variety of studies showing encouraging results for disease subtyping and clinical outcome prediction [43][44][45][46][47][48]. For example, the recent application of ML to tumour transcriptomes and methylomes uncovered molecular subgroups of hepatocellular carcinoma with highly distinct patient survival rates [49].…”