Over the summer, the MX Biobank Project research trainees participated in the “Closing the Genome Research Gap symposium: An international symposium on genomics and health“.
The symposium focused on the biased distribution towards European-decent populations in genomic research and the consequences in result interpretability and health outcomes that this haves, considering the vast amount of genetic variation that has gone understudied.
María José Palma and Andrés J. Kaufmann, traveled to McGill University in Montreal, Canada where they had lectures and discussions about statistical methods to study large and diverse cohorts, the health applications of such studies and the ethics and good practices in how to approach indigenous and vulnerable populations to perform genomic research.
Figure taken from: https://www.nature.com/news/genomics-is-failing-on-diversity-1.20759
Andrés J. Kaufmann had the opportunity to present a poster titled: “Imputation performance in Latin Americans: Expanding the imputation reference panel with Native American Genomes”. The basic idea of the poster was to discuss the fact that imputation performance in Native-American genetic variation is significantly worse than other ancestries and how this is due to underrepresentation of Native American genomes in current reference panels (Fig 1.A).
Then we showed that by expanding the current reference with publicly available Native American genomes we can improve performance (Fig 1.B). These results are important because imputation is a basic step in genetic research, by improving its performance in Latin American populations we can properly study the health consequences of the genetic variation in these populations.
Figure1. A) Imputation accuracy. Imputation accuracy by local ancestry and by Latin American Population, CLM(Colombia), PEL (Peru), MLX (Mexico) and PUR (Puerto Rico). B) Accuracy comparison between reference panels. Plot showing the comparison between runs of imputation using different references. Dashed lines represent the expanded reference.
María José Palma presented a poster titled: “The Mexican Biobank Project: Building Capacity for Big Data Science in Medical Genomics in Admixed Populations” in order to introduce the origin of the data related to MXBioBank project, its main goals as well as some preliminary results.
PCA. Principal Components Analysis showing the continental ancestry gradient of the MXB cohort (analysis conducted
using 1770 high quality genotyped individuals available as of May 2018).