Extended Data Fig. 4: Prediction scores for different accessible tissues as reference.
From: Hypergraph factorization for multi-tissue gene expression imputation
For each target tissue, we predicted the expression values based on accessible tissues (whole blood, skin sun exposed, skin not sun exposed, and adipose subcutaneous). We report the Pearson correlation coefficient between the predicted values and the actual gene expression values. For any given target tissue, we used the same set of individuals to evaluate performance, namely individuals in the validation and test sets with collected gene expression measurements in all the corresponding tissues. Target tissues represented by less than 25 test individuals were discarded. HYFA attains the best performance in 32 out of 38 tissues when all accessible tissues are simultaneously used as reference. Boxes show quartiles, centerlines correspond to the median, and whiskers depict the distribution range (1.5 times the interquartile range). Outliers outside of the whiskers are shown as distinct points. The top axis indicates the total number of samples for every target tissue.
