The overall accuracy of the method was 90%, as shown in the corresponding confusion plot (Supporting Information, Figure S1) and is similar in both the validation and testing datasets, meaning that there was no over fitting.
In their recent RNA-seq study of X A ratios, Xiong and colleagues [ 19] implemented what has been termed "unique" mapping parameters to map their datasets, meaning each short sequence read is aligned to the best position in the genome while any read which maps to multiple positions is excluded from the output.
For example, the activity types in ChEMBL, such as "IC50", "EC50", and "Potency", are not exposed in the RDF dataset, meaning that it is not yet possible to reason or group over similar activity types.
We report 100% accuracy for unsupervised classification in the Weizmann dataset, meaning that, in all the cross-validation folds, the 9 detected clusters actually coincide with the 9 different actions, and all the testing sequences are classified correctly.
However, the drawback of this approach is that the optimum parameters for spectral subtraction are empirically estimated from a development dataset, meaning that the late reverberation cannot be subtracted correctly as it is not precisely modeled.
There is no correlation of biopsy rate and cancer detection in this dataset, meaning that potentially there are many unnecessary biopsies performed.
Retrospective review of their ICU admissions for the study period identifies a further 311 patient episodes not entered in the dataset, meaning that 77% of episodes were captured across the ICCTG network.
Such a well-balanced performance is observed despite the highly unbalanced nature of the dataset, meaning that the number of subjects with the slow phenotype is almost an order of magnitude larger than that of subjects with the rapid phenotype.
There were 97 GO terms that were either enriched or underrepresented in the H. desmarestianus dataset, meaning there were either proportionally more or proportionally less genes (respectively) in the H. desmarestianus dataset with those GO terms than in the combined D. spectabilis and C. baileyi datasets.
Similarly, only a minority (22.8-65.022.8-65.0%s could be aligned onto the O. latipes UniGENE dataset, meaning that our O. melastigma RNA-Seq data comprises a significant portion of transcribed sequences that could not be unaligned and thus might be absent in the existing freshwater medaka genome and transcriptome sequences.
Similar classification AUCs to the cross-validated experiments indicate that the selected features are robust between datasets, having meaning even in other patient cohorts.
