An integrative analysis of genetic factors reveals dysregulation of cytokines, immune signaling, and T cell activity as the underlying immune mechanisms in autoimmune immune thrombocytopenia

Abstract

Background

Immune thrombocytopenia (ITP) is an autoimmune disorder characterized by immune-mediated platelet destruction, leading to an abnormally reduced platelet count. While numerous susceptibility genomic loci have been identified, the genetic mechanisms and pathways driving ITP remain poorly understood. This limits treatment options to broad immunosuppressants that increase patient vulnerability.

Objective

This study aims to uncover the functional and biological significance of ITP-associated genetic variations by integrating bioinformatics approaches. It seeks to identify functional SNPs, key immune pathways, and potential drug targets to enhance understanding of ITP pathogenesis and support the development of targeted therapies.

Methods

An integrative bioinformatics approach was employed to identify expression quantitative trait loci (eQTL) and pathogenic SNPs, reconstruct protein-protein interaction (PPI) networks, perform gene ontology analysis, and explore potential drug targets.

Results

The study identified 60 eQTL and 6 pathogenic SNPs associated with ITP, along with over 300 gene ontology processes. 14 hub genes in the PPI network were linked to key immune mechanisms, including T cell dysfunction (CD40, CTLA4, FOXP3, IL-10, IL-4, TBX21), cytokine dysregulation (IFNG, IL-6, IL-10, IL-4, TNF-α, TGFB1), JAK/STAT signaling (JAK2, STAT1, STAT3), and pattern-recognition (TLR4). TNF-α emerged as the top-ranked hub gene. Additionally, several platelet related genes (HPA2, MPL, PRKCA, PTPN11, and others) were implicated in the analysis.

Conclusion

Functional SNPs and hub genes identified in this study serve as potential biomarkers for ITP diagnosis and prognosis. Cytokine pathways and T cell subsets were highlighted as central players in ITP pathogenesis. The drug-gene interaction analysis further suggests potential therapeutic avenues through drug repurposing, offering insights into novel treatment strategies.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12959-026-00833-0.

Introduction

Immune thrombocytopenia (ITP) is an autoimmune bleeding disorder characterized by autoantibodies-mediated destruction of platelets, leading to severe reduction in the platelet count and insufficient platelet production [1]. Autoimmune disorders are typically regarded as rare, however, epidemiological research suggests a potential rise in the incidence and prevalence of ITP in the recent past. Although this increase may also be due to improved diagnosis, healthcare accessibility, and disease reporting, rather than a true shift in the disease incidence. The epidemiological data on ITP remains limited and variable across different countries with only small number of population-specific epidemiology studies been reported. For instance, a French study reported an incidence of 2.9/100,000 person-years [2] while a Danish study found that the 5-year incidence of primary ITP increased from 1.01 per 100,000 person-years in 1980–1984 to 5.71 in 2010–2014 [3]. Similarly, another retrospective cohort study reported an annual incidence of 6.1 per 100,000 persons from the year 2010–2016 [4]. This potential increase in the incidence and prevalence, combined with its multifactorial nature and unknown etiology, poses significant challenges in the clinical management of ITP.

Both genetic and environmental factors are widely accepted to play a role in the pathogenesis of autoimmune diseases, including ITP [5]. Several pathophysiological mechanisms are known to be involved in ITP pathogenesis. For example, an immune attack on the glycoproteins expressed by the platelets on their surface results in platelets’ phagocytosis in the liver and spleen [1]. Alternatively, a direct attack of the platelets by T cytotoxic cells leads to their eventual destruction resulting in a reduction in platelet counts and eventually ITP [1]. However, despite the continued efforts, the genetic mechanism remains largely unknown. Nonetheless, previous studies have identified and reported several Single Nucleotide Polymorphisms (SNPs) as the risk factors, critical genetic markers, and predictors for predisposition to ITP [5]. However, the biological function and the potential underlying mechanism of ITP pathogenesis involving these SNPs remain ambiguous. This necessitates the implementation of comprehensive gene expression and proteomics studies to understand the biological significance of these SNPs in ITP which are resource and time intensive.

In such a scenario, computational methods that provide important leads towards predicting molecular factors of disease pathogenesis will be highly instrumental. There are two main mechanisms by which a genetic variant can act on a phenotype - either through a direct change in the protein’s structure and subsequently its function or through an alteration in the regulation of gene expression and the encoded protein. Amino acid substitutions caused by genetic variations can be easily predicted using pathogenicity prediction tools [6, 7]. Alternatively, the discovery of expression quantitative trait loci (eQTLs) is instrumental in studying the association of SNPs with gene expression, offering insights into the molecular mechanisms underlying diseases and complex traits. The discovery of such genetic variants that explain variations in the gene expression levels holds great promise not only for characterizing functional sequence variation but also for the understanding of fundamental mechanisms of gene regulation and interpretation of genome-wide association (GWAS) studies. Hence, identification of eQTLs enables to prioritize causal variants and infer meaningful regulatory relationships that are often challenging to deduce from GWAS studies alone [8]. Integrating eQTL data with other predictive approaches, such as chromatin accessibility, transcription factor binding studies, impact of amino acid substitution on protein functions, and protein-protein interactions (PPI) network analysis, allows for a more comprehensive understanding of the functional implications of the genetic variants.

Recent advancements in the eQTL resources have provided extensive data on tissue-specific regulatory mechanisms, further enhancing our understanding of the gene regulations across different human tissues. Besides these resources, the comprehensive databases, like RegulomeDB, can analyse a given variant by mapping its genomic coordinates to functionally active regions identified through functional genomic assays such as Transcription Factor ChIP-seq (TF ChIP-seq) and DNase-seq as well as those from footprints and QTL data [9]. Additionally, the functions of variant-associated genes can be explored using the functional annotation databases [10], alongwith the analysis of gene interrelationships through PPI networks [11].

The growing availability of such computational resources has enhanced the ability to draw meaningful insights from complex genomic datasets, paving the way for improved diagnostics, therapeutic targeting, and precision medicine strategies. Benefiting from these bioinformatics advancements, numerous studies have showcased the utility of integrating meta-analysis and bioinformatics pipelines to prioritize SNPs and associated genes for specific disease [12]. Therefore, by leveraging a similar bioinformatics strategy, this study aims to bridge the gap between genotype and phenotype, and improve our understanding of etiology and therapeutic targets for ITP. In the present study, a computational approach is employed to build an understanding of the functionality and potential role of SNPs in ITP etiopathogenesis. Further, an integrated data-driven approach is used to uncover the biological functions and potential drug targets for the ITP-related genes and SNPs.

Methods

The workflow of the complete approach is shown in Fig. 1.

Fig. 1.

Schematic of the workflow for the functional analysis of SNP, genes, and therapeutic targets for ITP

Literature search for meta-analysis

For the meta-analysis, PubMed, Google Scholar, and Web of Science were used to retrieve literature using keywords, such as “ITP”, “immune thrombocytopenia”, “SNP”, “single nucleotide polymorphism”, “idiopathic immune thrombocytopenia”, “immune thrombocytopenic purpura”. Further, the eligibility of studies was checked using several strategies: (i) duplicate articles from the database search were removed using DOIs, (ii) the titles and abstracts were referred to find out the “article type” (i.e., research, review, book chapter, editorial, and others) and SNP studies from the list, (iii) only research articles with SNP studies were considered which were further screened for reporting a statistical correlation between the SNPs and ITP. The resulting research articles were thoroughly studied to prepare a list of ITP-associated SNPs, genes, country/ethnicity, sample type, cohort size, genotyping method, and DOI for each study.

SNP mapping

The ITP linked SNPs were mapped onto the RegulomeDB [9] to identify the eQTL SNPs. For this, entries of studies reporting ITP correlated SNPs were subjected to deduplication, and the SNPs lacking rsID information were excluded, as RegulomeDB analyzed only the rsIDs as input. Further, pathogenicity prediction for SNPs was performed using PolyPhen-2 [6] and SIFT (Sorting Intolerant from Tolerant) [7] to predict the functional impact of SNPs on the protein structure and function.

Biological network building and visualization

Using STRING (Search Tool for the Retrieval of Interacting Genes) [11], a PPI network was generated for the ITP genes (those genes which had ITP-related SNPs) using default parameters (interaction score = 0.4). This PPI network was further visualized and the network centrality was evaluated to identify the hub genes using cytoHubba plugin of the Cytoscape tool [13, 14]. For visualization, default STRING parameters were adjusted as follows: minimum interaction score set to 0.9 (highest confidence), edge thickness scaled to reflect strength of supporting evidence, and nodes were colored red (hub genes) or green (non-hub genes).

Biological functions of the ITP gene set

The ITP gene set was classified and annotated using Enrichr web server to understand the biological functions [10]. For this, the three classes of Gene Ontology (GO) terms were analysed for these genes, namely GO Biological Process, GO Molecular Function, GO Cellular Compartment. Stringent parameters (adj p-value ≤ 0.05 and gene count > 3) were employed to shortlist the gene sets and terms with strong enrichment evidence.

Identification of drug targets for hub genes

To identify the drug-gene interactions among the hub genes, the Drug-Gene Interaction Database (DGIdb) was queried [15] and the FDA-approved drugs were prioritised for subsequent analysis. The adverse effects associated with the potential drugs were looked for by utilizing the SIDER tool [16].

Results

Following the comprehensive literature review analysis, around 250 articles were screened, of which 148 research articles reporting SNP studies were retrieved. Of these, 109 articles were included for downstream analysis, as these reported a statistical correlation between the SNPs and ITP (Supplementary Table 1).

Biological implication of ITP SNPs

For 89 ITP-associated SNPs, eQTL analysis was done to assess the associations between these SNPs and the gene expression levels (Supplementary Table 2). Consequently, 60 SNPs with a high probability score (score 1a to 1f) were found to have high potential to act as regulatory variants (Supplementary Table 2). Further, 6 SNPs were predicted to be likely pathogenic/damaging by PolyPhen-2 (Supplementary Table 3A) and SIFT (Supplementary Table 3B). These SNPs are associated with the genes ABCB1, IL-27, LAG3, MPL, MUC20, and PDCD1. Out of these, one SNP with rs3762739 (MUC20 gene) was predicted by both PolyPhen-2 and SIFT.

Protein-protein interaction network analysis

The inter-relations between the proteins can reflect their biological roles in disease development, hence the network topology of the PPI networks (Fig. 2) was analysed based on three algorithms, namely betweenness, closeness, and degree, and the top 20 genes from each were shortlisted (Supplementary Table 4). When the top 20 ranked genes of all the three centrality parameters were compared, 14 common genes were found and TNF-α was consistently ranked top amongst all. These 14 genes were selected as the hub genes, which included CD40, CTLA4, FOXP3, IFNG, IL-10, IL-4, IL-6, JAK2, STAT1, STAT3, TBX21, TGFB1, TLR4, and TNF-α. Of these, CD40, CTLA4, FOXP3, IL-10, STAT1, and TNF-α had eQTL SNPs as shown in the SNP-gene relationship investigation (Sect. "Biological implication of ITP SNPs").

Fig. 2.

Schematic illustration of a protein-protein interaction network (PPI) of ITP genes constructed using the STRING database. For visualization, the default STRING parameters were adjusted: interaction score = 0.9, edge thickness reflects confidence, node colors - red (hub genes) or green (non-hub genes)

Pathway enrichment

While mapping the ITP genes onto their biological pathways, 327 Gene Ontology Biological Process (GOBP) (Supplementary Table 5), 23 Gene Ontology Molecular Function (GOMF) (Supplementary Table 6), and 4 Gene Ontology Cellular Compartment (Supplementary Table 7) were found. All the significant pathways and functions (adj p-value ≤ 0.05 and gene count > 3) were manually curated and categorized into broad biological functional classes as shown in Fig. 3. The top 20 biological processes (based on adjusted P-value) are primarily involved in cytokines regulation (IL-6, IL-10, and TNF-α). Several other immune pathways were also observed, including inflammatory responses, cellular responses to bacteria, viruses, lipopolysaccharides, and lymphocyte proliferation (Supplementary Table 5). In addition, other pathways related to signaling, kinases, transcriptional activity, metabolism, developmental processes, and platelet activation were also identified amongst the highly significant (adj p-value ≤ 0.05 and gene count > 3). Further, the top 20 molecular functions included mainly the processes related to cytokine, kinase, phosphatase, and protease activities (Supplementary Table 6).

Fig. 3.

Gene ontology (GO)-based functional classification of ITP genes (a) GO Biological Process, (b) GO Molecular functions, (c) GO Cellular Compartments

Drug-gene interactions and drug side effects

For the 14 hub genes, a total of 464 gene-drug interactions, comprising 172 FDA approved drugs, were obtained from DGIdb database (Supplementary Table 8A). The FDA approved drugs candidates were assessed for their safety profile using SIDER Side Effect Resource. For this, we focused on side effects classified as “very common”, that is, mild and non-severe. Among the 172 drug candidates, 20 had reported side effects in the meddra category “very common” (Fig. 4, Supplementary Table 8B).

Fig. 4.

Potential FDA approved drugs targeting hub genes and exhibiting mild and non-severe side effects (Blue boxes represent hub genes, green boxes represent in-silico predicted drugs and having published evidence in context to ITP, transparent boxes represent in-silico predicted drugs)

Discussion

This study leverages the published literature to prioritize ITP-associated SNPs and genes, revealing their potential biological functionality using a comprehensive data-driven approach. The functional analysis overwhelmingly showed that the genes are involved in immunomodulatory processes (e.g., cytokines, immune signaling, T cell activity), which remain predominant in the GO and hub gene analyses. Although several pathways related to signaling and kinase activity, transcriptional regulation, metabolism, and platelet activation were observed, their relatively lower enrichment strength (i.e., gene count) suggests that immune dysregulation is the primary process driving ITP pathophysiology. The 14 essential hub genes are functionally interlinked by either acting antagonistically or synergistically in the immune dysfunction responsible for ITP pathophysiology (Fig. 5).

Fig. 5.

Potential immune mechanisms mediated by the hub genes in ITP etiopathogenesis. Figure is created using Biorender

Immune dysfunction of T cells activity in ITP

Our analysis revealed six hub genes that are intricately associated with the modulation of T-cell immune responses, namely CD40, CTLA4, FOXP3, IL-10, IL-4, and TBX21. The costimulatory CD40 typically promotes activation of autoreactive B cells, while CTLA4 co-inhibits T-cell activation, that is, dampens the CD40-mediated activation signals. In accordance, ITP patients initially showed reduced CTLA4 levels which increased after dexamethasone therapy [17], while CD40L levels were found to be elevated [18]. This means, CTLA4 may protect against excessive immune response in ITP. On the other hand, antiplatelet antibodies-mediated ITP could possibly involve CD40-CD40L co-stimulation [19] and therefore, blockade of CD40-CD40L interaction could be a therapeutic option for suppressing ITP [20]. Further, about > 95% of the circulating soluble CD40L is derived from the platelets [21] and it can initiate inflammatory responses through T and B lymphocytes [22].

A dysfunctional regulatory T cell (Tregs) FOXP3 leads to severe autoimmune disorders, including ITP [23]. Tregs produce anti-inflammatory cytokines, such as IL-10 and TGF-β1, which downregulate pro-inflammatory signals and maintain immune homeostasis [24]. The FOXP3⁺ Tregs and APCs interaction via CTLA4-B7 leads to anti-inflammatory cytokines production, including TGF-β and IL-10 [25]. In line with this, notable decrease in Tregs frequency [26], FOXP3 expression, and serum levels of TGF-β1, IL-4, and IL-10 is observed in ITP patients [27]. Additionally, severe ITP cases show significant reduction in CD4 + CD25 + Tregs [26]. As the inhibitory effect of Tregs towards the effector T cells (Teffs) is largely exerted by IL-10, insufficient secretion of IL-10 compromises this inhibitory capability in new ITP patients [28]. Yet the regulatory role of IL-10 remains highly contradictory as its expression may increase [29] or decrease [30] in ITP. An increase in IL-10 levels could be related to macrophages activation, which have been reported to be stimulated in ITP patients by platelet autoantigen, leading to T cell activation [29]. This suggests that disruptions in Tregs activity, FOXP3, TGF-β1, IL-4, and IL-10 contribute to the dysfunction of inhibitory immune responses in ITP.

The data on polarization of the immune system towards certain cytokines is limited and contradictory. For instance, expression related discrepancies are known for cytokines including IL-4 [27, 31] and IL-10 [29, 30] in the ITP population, making it difficult to characterize their inflammatory roles in ITP pathogenesis. Such contradictory findings suggest that a highly multifaceted and convoluted immune network or mechanism might contribute to ITP.

Though the role of TBX21 is not known in ITP, as a transcription factor it can bind to IFNG promoter to induce its transcription, and in turn drive the Th1 effector cells differentiation [32].

Inflammation and cytokine dysregulation

The hub genes also included several other cytokines, including IFNG, IL-6, TNF-α, TGFB1, IL-4 and IL-10 (as discussed in Sect. "Immune dysfunction of T cells activity in ITP"). The proinflammatory IFNG, stimulates pathogenic Th1 polarization, through JAK/STAT activation, and inhibits regulatory Th2 cells. On the contrary, IL-4 drives the development of naive Th cells into the Th2 subset which regulates antibody production, counterbalancing IFNG driven Th1 responses. Consistently, ITP patients exhibit significantly higher IFNG expression and remarkably lower IL-4 expression [33]. Further, a reduced TGF-β1 expression correlated with IFNG elevation in chronic pediatric ITP patients [34]. Additionally, platelet derived TGF-β1 induces expansion and reprogramming of immunosuppressive myeloid-derived suppressor cells (MDSCs), implying that platelet recovery not only serves as an endpoint of remission, but also induces immune tolerance in ITP [35].

In our network topology analysis, pro-inflammatory TNF-α was ranked top, indicating it to be the master immune regulator in ITP. Under in-vitro conditions, TNF neutralization in the T cell-monocyte co-culture enhanced the expansion of pre-existing ITP Tregs more than in healthy controls, suggesting that TNF-α inhibits Tregs in ITP [36]. Corresponding to this, increased levels of TNF-α were observed in ITP patients [30], and hence TNF-α blockade is being explored as a therapeutic strategy for the management of ITP patients [36].

IL-6 also promotes pro-inflammatory T cells (Th17 and Teff) and suppresses FOXP3 Tregs through JAK/STAT pathway signaling [37]. Further, in combination with TGF-β, it promotes pathogenic Th17 cells [38]. Also, a significantly higher IL-6 expression has been reported in ITP patients, further supporting its role in the pathogenesis of the disease [29, 31].

JAK/STAT-mediated signaling disrupts immune homeostasis in ITP

JAK2, STAT1, and STAT2 are a part of the JAK/STAT signaling pathway wherein the STAT molecules are activated by several ligands, such as IFNG, IL-6, and others. An aberrant phosphorylation of STAT3 has been observed in the ITP patients [39] and it is known to regulate the pro-inflammatory Th17 cells such that its inhibition in CD4 + T cells prevents autoimmunity [40]. Further, an IL-6 mediated signaling promotes phosphorylation and activation of STAT3 [41] and it is reported that ITP patients have higher IL-6 levels than healthy controls [29]. It is interesting to note that several of the hub genes are functionally linked with the JAK/STAT pathway, making it a potential candidate in ITP treatment. Though the direct role of JAK molecule in ITP is unexplored, it remains an indispensable part of the STAT-mediated signaling, and therefore may have a mechanistic role in ITP pathogenesis.

Roles of the pattern-recognition receptor TLR4

TLR-mediated pathogen recognition drives the differentiation of naive T cells into Th1 and Th2 subsets in a cytokine-dependent manner. Studies have shown that a vigorous Th1 response leads to autoimmunity, the Th1/Th2 balance determines the outcomes of immune response in immune disorders and TLRs are critical for this [42]. In line with this, an increased level of TLR4 expression in monocytes of ITP patients was observed, suggesting it may play a role through the TLR4-cytokine-CD4⁺ Th cell pathway [43]. Further, Cheng et al. reported that increased TLR2 and TLR4 expressions in ITP patients may contribute to the pathophysiological progression of the disease by increasing the secretion of TNF-α and IFNG [44]. Conversely, decreased TLR4 levels in monocytes appears to cause Tregs abnormality in ITP, since the Th17/Treg cell ratio decreased with lipopolysaccharides (LPS) stimulation of TLR4 and increased with anti-TLR4 treatment in ITP patients [45]. The same study also reported that IL-10 and TGFB1 levels increased with LPS and decreased with anti-TLR4 [45]. Besides, TLR4 present on the platelets and megakaryocytes can induce platelet adhesion and aggregation in LPS-induced thrombocytopenia [46], however the mechanistic understanding of its role in ITP needs further explorations.

ITP-associated genes affect platelet production

Several of the genes having SNP association with ITP are expressed by platelets, and hence have a direct link with them. These include CBL, DNAM1, eNOS, FCGR2A, HPA2, HPA5, IL18, IL1β, MPL, NLRP3, PRKCA, PTPN11, PTPN22, SIRT1, TIMP3, and TUBB1. Amongst these, MPL (myeloproliferative leukemia virus oncogene) stands out due to a predicted high-impact SNP (potentially deleterious), suggesting it may be a critical gene in the pathogenesis of ITP, and therefore further experimental investigation could be targeted in this direction (Supplementary Fig. 1). As MPL interacts with thrombopoietin (TPO) to regulate platelet production, its expression on megakaryocytes and platelets is essential to prevent megakaryocytosis and myeloproliferation [47]. Strikingly, drugs like Eltrombopag and Romiplostim, which have revolutionized ITP treatment by effectively improving platelet counts, are thrombopoietin receptor agonists (TPO-RA) that act via various signaling pathways including JAK/STAT [48]. Given the role MPL plays in this process, it would be highly valuable to investigate the relationship between TPO-RA therapy and MPL-associated SNPs, particularly in TPO-RA refractory ITP cases.

We also identified some important overlaps in our findings such as genes having eQTL SNPs as well as pathogenic SNPs (ABCB1, IL-27, LAG3, PDCD1) and those carrying pathogenic SNPs as well as having direct role in platelet production (MPL) (Supplementary Table 3A, Supplementary Fig. 1). Of these, SNP rs3762739 linked with MUC20 gene is particularly interesting mainly because it was predicted to be potentially pathogenic by multiple computational tools, underscoring its possible relevance in ITP. Such overlaps can reflect the key genes which can be further prioritized as potential biomarkers for ITP.

Therapeutic alternatives via drug repurposing in ITP

ITP treatment approaches have been evolving since 11th century [49]. Yet currently immunosuppressive corticosteroids remain first-line of therapy, while the non-responders are subjected to TPO-RAs or other alternatives [50]. The reliance on non-targeted immunosuppressive therapies often causes increased vulnerability in patients. While newer, targeted drugs are in development, clinical trials take time. Therefore, drug repurposing approaches can offer a faster, safer, and cost-effective alternative.

Our gene-drug interaction predicted that 8 out of the 14 hub genes (TGFB1, IL-10, TNF, JAK2, IFNG, IL-6, TLR4, and CTLA4) were linked with FDA-approved drugs with predominantly non-severe and mild side-effect profiles, making them potential candidates for repurposing. Several drugs predicted in our analysis are supported by literature evidence for ITP, namely Imatinib [51], Lenalidomide [52], Ruxolitinib [53], and Thalidomide [54]. These convergences between our in-silico predictions and real-world therapeutic responses strengthens our findings, and suggest that targeting pathways involving JAK2, TNF-α, and CTLA4 (showing interaction with the four drugs) may offer effective immunomodulation in ITP management. Although the drug-gene interaction analysis was filtered using SIDER profiles to exclude drugs with serious adverse effects, these findings need to be validated and clinical trials must be conducted prior to their therapeutic recommendations for ITP. Nonetheless, the overlap between predicted targets and literature emphasizes the potential for drug repurposing in ITP.

Finally, as literature on ITP-associated SNPs shows considerable heterogeneity, in terms of the ethnicity, cohort size, sample type, and genotyping methods employed, some SNPs might appear significant in one cohort but not in others. These limitations highlight the need for larger, ethnically diverse studies using standardized methods to reliably validate genetic variants implicated in ITP.

Conclusion

This study presents a comprehensive computational framework to integrate the existing genetic findings and predict the potential functional consequences, thereby reconfirming the known biomarkers and offering valuable insights into the putative genetic and molecular basis of ITP. However, additional experiments are required to confirm their mechanistic and translational context as potential diagnostic or prognostic biomarkers. Further, the study provides important leads towards more effective, targeted, and safer alternative therapies by overcoming the challenges of the traditional drug development process, including tedious drug discovery and cost, resource, and time intensive in-vitro and in-vivo validations followed by clinical trials. Despite several advantages, this study has some limitations, including the small dataset, primarily due to the limited number of SNP studies on ITP and the exclusion of studies lacking statistically significant SNP-ITP correlations. The lack of correlation in some studies could be attributed to the small sample sizes. Nonetheless, these limitations highlight the need for GWAS studies in large ITP cohorts in the future. Additionally, since SNP and eQTL data were sourced from different cohorts, validating eQTL roles within the same patient populations using a multi-OMICS approach is essential for future research. Besides, the damaging effects identified for certain SNPs at protein-level, are based on computational analyses, and therefore require further experimental validation to determine their actual biological roles in disease pathogenesis. Further, the drug-gene interaction analysis intends to propose exploratory drug candidates, therefore these findings must be validated experimentally to assess their mode of action, effects on disease pathogenesis and severity, and potential side effects in ITP.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

APC

Antigen Presenting Cells

DGIdb

Drug-Gene Interaction Database

eQTL

Expression Quantitative Loci

GO

Gene Ontology

GOBP

Gene Ontology Biological Process

GOMF

Gene Ontology Molecular Function

GWAS

Genome-Wide Association Study

ITP

Immune Thrombocytopenia

LPS

Lipopolysaccharides

MDSCs

Myeloid-Derived Suppressor Cells

MPL

Myeloproliferative Leukemia Virus Oncogene

PPI

Protein-Protein Interaction

SIDER

Side Effect Resource

SIFT

Sorting Intolerant from Tolerant

SNP

Single Nucleotide Polymorphism

STRING

Search Tool for the Retrieval of Interacting Genes

TF ChIP-seq

Transcription Factor ChIP-seq

TPO-RA

Thrombopoietin Receptor Agonists

Tregs

Regulatory T Cells

Author contributions

Pratyusha Patidar: Conceptualization, Investigation, Methodology, Data curation, Formal analysis, Visualization, Writing - original draft, Writing - review & editing. Tulika Prakash: Overall supervision, Conceptualization, Project administration, Methodology, Resources, Writing - review & editing.

Funding

None.

Data availability

All data generated or analysed during this study are included in this published article (and its supplementary information files).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.