missRows

The missRows software tool proposes a multiple imputation approach in a multivariate framework to deal with missing row values in incomplete datasets. Missing row values are challenging because most statistical methods cannot be directly applied to incomplete datasets. This tool focuses on multiple-factor analysis (MFA) to compare and integrate multiple layers of information. The method fills the missing rows with plausible values, resulting in M completed datasets. MFA is then applied to each completed dataset to produce M different configurations (the matrices of coordinates of individuals). Finally, the M configurations are combined to yield a single consensus solution.

The authors assessed the approach on two real omics datasets with different patterns of missingness. They compared it with two other approaches: regularized iterative MFA (RI-MFA) and mean variable imputation (MVI-MFA). The solution's suitability was determined against the true MFA configuration obtained from the original data for each configuration resulting from these three strategies. A comprehensive graphical comparison was produced showing how the MI-, RI- or MVI-MFA configurations diverge from the true configuration. Additionally, two approaches, namely confidence ellipses and convex hulls, were described to visualize and assess the uncertainty due to missing values.

The tool provides a useful and attractive method for estimating the coordinates of individuals on the first MFA components despite missing rows. It considers the uncertainty of MI-MFA configurations induced by the missing rows, thereby allowing the reliability of the results to be evaluated. The MI-MFA configurations were close to the true configuration even when many individuals were missing in several data tables. This method considers the uncertainty of MI-MFA configurations induced by the missing rows, thereby evaluating the results' reliability.