Van Deuren, V. | bioRxiv Science

bioRxiv2024

Bridging immunotypes and enterotypes using a systems immunology approach

Unveiling the systemic effects of disease and health requires an holistic approach that has mainly revolved around well established, directly determinable molecular relationships such as the protein synthesis cascade and epigenetic mechanisms. In this study, involving 394 individuals, we found direct linkage of branches spanning human biological functions often not studied in conjunction, using clinical data, gut microbial abundances, blood immune cell repertoires, blood transcriptomic and blood T cell receptor data. Contrary to current paradigms, we demonstrate that immunotypes and enterotypes are orthogonal, likely fulfilling distinct roles in maintaining homeostasis, only bridged via the blood transcriptome. We also identified two distinct inflammatory profiles: the first driven by interferon signalling and the other characterised by non-viral, NF-kB and IL-6 markers. Lastly, we present compelling data showing strong associations of the Ruminococcaceae and Christensenellaceae bacteria with a healthy immunotype and transcriptomic pattern, highlighting their potential role in immune health. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=194 HEIGHT=200 SRC="FIGDIR/small/625344v1_ufig1.gif" ALT="Figure 1"> View larger version (52K): org.highwire.dtl.DTLVardef@c6b5eorg.highwire.dtl.DTLVardef@15d0c3forg.highwire.dtl.DTLVardef@1cf29eeorg.highwire.dtl.DTLVardef@1c9ff8d_HPS_FORMAT_FIGEXP M_FIG C_FIG

systems biology↗

bioRxiv2024

Revealing the hidden sequence distribution of epitope-specific TCR repertoires and its influence on machine learning model performance

Numerous efforts have been made to decipher the epitope-T cell receptor (TCR) recognition code. Both simple machine learning techniques and deep learning strategies have been used to train models to predict the binding of epitopes by TCR sequences. A good training data set rests at the basis of every accurate prediction model, yet little attention has been given to the composition of these data sets. In this paper, we studied the natural distribution of TCR sequences within epitope-specific TCR repertoires, i.e. a set of TCRs binding the same epitope, and its impact on the predictability of TCR-epitope interactions. We found that the observed diversity of these repertoires can result from a smaller set of core binding motifs constrained by TCR generation. Moreover, a clear relationship was found between the sequence distribution of the training data and performance metrics, emphasizing the importance of the used ground-truth data when using machine learning models in this domain. Taken together, these findings inform data set composition to help push epitope-TCR prediction models to the next level.

bioinformatics↗