Transcriptomic data provide a systematic multivariate snapshot measurement of the state of a biological system. In my team, we develop transcriptomics data integration and modelling frameworks to reconstruct systems dynamics, identify and precisely quantify hidden phenotypic variation at multiple scales, from the molecular level to organismal physiology, and to make quantitative predictions about genetic and non-genetic cellular and molecular and mechanisms of phenotypic variation including at single-cell and single-individual level. In contrast to others, our strategies privilege interpretability and robustness. I will present examples of how I applied these strategies to predict complex phenotypes and mechanisms at the single-cell and single-organism level.  In particular, I will present RAPToR, a simple but precise computational method to infer absolute physiological age of a system from its transcriptome exploiting existing time series data as reference1. Importantly, ages estimated on the same reference are comparable across conditions, genetic backgrounds even across species. RAPToR, works in multiple model organisms or humans for both development and aging, for bulk, dissected tissues, single-individual and single-cell data. Moreover, provided with tissue/process -specific annotation RAPToR can provide tissue specific age estimates from whole-animal data.  Estimated physiological ages, (in combination with chronological age) can be used not only to precisely quantify the effect (and time of action) of genetic, environmental and inter-individual variation on development or aging speed but also to quantify their specific effects of on gene expression, suggesting their specific molecular determinants. In summary, large scale transcriptomic data can be exploited to both quantify complex phenotypes and learn molecular and cellular determinants of these phenotypes. These quantities inferred from gene expression can then be used to parametrize mathematical models to predict phenotypes.

References

1.Bulteau, R. & Francesconi, M. Real age prediction from the transcriptome with RAPToR. Nat Methods 1–7 (2022) doi:10.1038/s41592-022-01540-0

If you would like to attend the seminar, please register here.