We compared RIC predictions from the model trained on proteins with experimental functions to RIC predictions based on electronic annotation data.
(c) ROC plot for model trained on one genome and test on another.
Two predictions were made, one with the BN+SVM model trained on the full data set (prediction 1, or P1), and one with the BN+SVM model trained on the trimmed data set (P2).
Here cross-database segmentation refers to the segmentation of fingerprint images from one subdatabase using a classification model trained on another subdatabase.
(d) Predicting stitched SEs using model trained on constituents.
The model, trained on the mean fire data between 2000 and 2017, was able to explain 70% of total deviance.
We then applied a likely voter model trained on previous election cycles to create estimates for the 2018 electorate. 1.
We also tested whether a model trained on constituent data can predict the stitched regions and vice versa.
For further validation, a separate model, trained on 2006 2011 mean data, was used to predict burned area for 2001 2006 and 2011 2016 mean data.
It was particularly interesting to see that a model trained on genomic features, including GC content, phastCons, and repeat fraction, achieved AUC = 0.81.
The model trained on 1978 1984 was then projected onto annual climatologies for SIC, SST, and bare-rock locations from 1981 to 2010.
