Multi-view learning and quantitative genetics to identify the molecular basis of adaptation to chemical pollutants

Descripción del proyecto

Adaptación de las poblaciones naturales a los plaguicidas

El uso de plaguicidas en la agricultura es una práctica habitual, pero poco se sabe sobre los efectos que provocan a largo plazo sobre las especies que no son su objetivo, como el zooplancton de agua dulce. El objetivo del proyecto MultiOmicsTox, financiado con fondos europeos, es investigar cómo la presencia prolongada de sustancias químicas en el medio ambiente durante numerosas generaciones altera la variación genética y afecta a la evolución molecular. Los investigadores estudiarán la evolución de «Daphnia» durante un período de ciento veinte años y asociarán los cambios moleculares con la exposición a plaguicidas. Así, obtendrán importante información sobre la verdadera vulnerabilidad de las poblaciones ante los plaguicidas y sus respuestas adaptativas. A largo plazo, estos datos contribuirán a la formulación de políticas medioambientales y sanitarias adecuadas.

Objetivo

My project proposes to understand what genetic and functional genomic variation contribute to the process of adaptation and to the evolutionary fate of natural populations when confronted with modern threats, such as multi-generational exposure to a chemical pollutant in the environment. The current environmental health policies and regulatory decisions are based on ad hoc methods and do not reflect true population susceptibility. My solution is to apply multi-view machine learning, combined with quantitative genetics, to analyse a huge volume of multi-omics data to advance Precision Toxicology that brings greater certainty in the causal links between chemicals and their adverse effects. My project focuses on the multi-generational effect of pesticides in shaping genetic variation and the molecular evolutionary trajectory of Daphnia obtained from resurrected subpopulations from within dated lake sediments spanning 120 years. The adaptive phenotypes at different doses of pesticides were scored in common garden experiments and samples were taken to produce associated multi-omics data (genomes, transcriptomes, regulomes and metabolomes). I propose utilizing this data to meet the following two objectives:(1) To use quantitative genetics for the determination of genetic susceptibility of the subpopulation to pesticide exposure; (2) To identify the mechanisms and forms of evolution that result in adaptation, by integrating multi-omics data using multi-view machine learning. Expected outcomes of this work will (a) fill a gap in mechanistic understanding of the adaptive responses of natural populations, (b) identify segregating genetic variation within genomes that regulates the pace and magnitude of an adaptive response to chemical pollutants, and (c) discover putative biomarkers that estimate exposure-related genetic susceptibility of populations to the multi-generational harmful effects of chemicals for setting site-specific controls on chemical pollutants.