Avocado: a multi-scale deep tensor factorization method learns a latent representation of the human epigenome

@iskander You should have a few weeks on Avocado, obviously, as the ~first~ foundation model for genomics. 🥲 https://t.co/p02mPunPto

@iskander TargetFinder paper is still cited in positive way and dataset from it is used in newer papers (e.g, https://t.co/o8DIbYBTG9)

At the end of the first round, teams were compared to two baselines: the average activity (see: https://t.co/04t5evPtwi) and Avocado (https://t.co/p02mPuESvo). Although many outperformed the average activity, the teams still had more work to do (a lower sc

@TiffanyTimbers @michaelhoffman @jtleek @choldgraf Somewhat similar to Virtual ChIP-seq, functional genomics imputation is a massive data problem. https://t.co/FuEifCYyDp hosts >10k experiments that can be turned into an incomplete (cell type, assay, g