Detecting complex, higher-order relationships in clinical datasets—especially for multifactorial conditions such as multiple sclerosis (MS)—is challenging with standard statistical approaches that compare group averages and may discard distributional information. The authors applied Genomic Informational Field Theory (GIFT), a method that ranks individuals on linear measures (for example, immunoglobulin titres) to retain and interrogate distributional structure. The study aimed to test whether GIFT can reveal humoral immune response associations with herpesviruses that are sex-dependent and that conventional statistics may miss.
The analysis used immunoglobulin titre data from 200 MS patients and 137 healthy controls. Humoral responses to four highly prevalent human herpesviruses were assessed to evaluate associations with MS susceptibility. The study received ethical approvals from the Ethical Committee of San Luigi Gonzaga University Hospital (CRESM Biobank approval, ref 18390/2019) and Nottingham Research Ethics Committee 2 (Ref. 08H0408-167). Samples were handled following UK Human Tissue Authority guidelines and all participants provided written informed consent.
The authors compared the performance of GIFT against conventional statistical frameworks applied to the same immunoglobulin dataset. Conventional approaches typically focus on average differences between groups and may not capture subtler, distribution-based signals or subgroups within the population. GIFT operates by ranking individual measurements to preserve information about the underlying distribution of datapoints and thereby aims to detect higher-order or subgroup-specific associations.
Using both analytical approaches, the well-established link between Epstein–Barr virus (EBV)—specifically antibodies against the EBV nuclear antigen EBNA1—and MS susceptibility was validated in both sexes. This replication supports the dataset and assay measurements and confirms that GIFT can detect known, robust virus–MS associations.
Beyond the EBNA1 replication, GIFT identified additional associations that were not observed with conventional statistical tools in this dataset. In female participants, GIFT linked MS susceptibility to humoral responses against herpes simplex virus (HSV), varicella zoster virus (VZV), and the EBV viral capsid antigen VCA. In male participants, GIFT associated MS susceptibility solely with cytomegalovirus (CMV) immunoglobulin levels. According to the authors, none of these sex-specific associations were detected using the conventional statistics applied to the same data.
The authors argue that conventional statistical tests, which typically compare group means, can lose information about the full distribution of individual measures and thus fail to identify subgroup-specific signals. By ranking individuals and preserving distributional relationships, GIFT can reveal complex phenotype–trait associations and identify subgroups within populations that may drive disease susceptibility signals. The study demonstrates GIFT’s capacity to both confirm a known association (EBNA1–EBV) and uncover additional sex-specific herpesvirus links to MS that remained undetected by average-based methods.
All data produced in the study are available upon reasonable request to the authors and require completion of a signed data transfer agreement. Funding sources declared include the MRC (CARP MR/T024402/1) and Associazione Italiana Sclerosi Multipla (2020 R-Multi 050). The authors declared no competing interests. The manuscript is posted as a medRxiv preprint on August 6, 2026, and has not been certified by peer review; the authors note that the work should not be used to guide clinical practice pending formal peer review.
The report highlights methodologic advantages of GIFT in detecting distributional and subgroup associations in immunoglobulin data, but as a preprint the findings have not yet been peer reviewed. Specific methodological details beyond the conceptual description (for example, algorithmic parameters, exact statistical comparisons, or effect sizes) are not reported in the source summary provided here. The authors propose that GIFT can be useful to detect complex phenotype–trait associations and to reveal underlying subgroups within populations, suggesting further work to replicate these sex-specific herpesvirus–MS links and to examine their biological and clinical relevance.