To improve the accuracy of prediction, we built a hierarchical identification pipeline based on nine prediction programs (Table 1).
A three-layer hierarchical identification rule based on nine prediction programs was used to identify putative secretory proteins in 158 fungal/oomycete genomes (208,883 proteins, 15.21% of the total proteome).
To support the recognition of patients at high risk for HL after BM, Koomen et al. developed a clinical prediction model based on five predictors, including: duration of symptoms prior to admission longer than two days, the absence of petechiae, cerebrospinal fluid (CSF) glucose level ≤0.6 mmol/L, Streptococcus pneumoniae as causative pathogen and the presence of ataxia during the illness [ 3].
From this dataset, we estimated safe winter chill for each point in time, with future predictions based on three IPCC greenhouse gas emissions scenarios (B1, A1B and A2), and averaged over three global climate models (CSIRO-MK3, HadCM3 and MIROC).
We used an extensive cross-validation process to define an applicability domain for model predictions based on two quantitative measures: prediction confidence and domain extrapolation.
A simplified version of the clinical prediction score based on three dichotomized abnormal laboratory results (blood albumin, sodium and urea, available for 7,608 patients, 92.3%) was developed.
Remarkably, the best prediction system based on three descriptors yielded a validated accuracy of 84.6 %.
There are far fewer matches to consider if we base our analysis on parsimony reconstruction of ancestral labels in the species tree, because parsimony-based reconstructions for an inner node are based on two or more predictions for the leaves (the extant species).
A general toxicity prediction can be based on three consecutive steps [2]: (1).
