To that end, we used a dataset divided into two parts, one for training and the other for ANN testing.
Then, for each run we computed the Weak Scaling Efficiency (WSE), that is the running time for one processing element (20 cores in our case) to process one work unit (1/4 of the dataset), divided by the running time for N processing elements to process N work units ((N=1,2 ldots 4) in our case).
The Mean Ranking is basically the sum of ranks obtained by a method in each dataset divided by the total number of datasets, as below: mathit{MR} = frac{sum_{j=1}^{mathit{nd}}mathit{ri}}{mathit{nd}} where nd is the number of datasets and ri is the rank of the method for dataset i.
This z-score uses the protein abundance value subtracted by the median value of the 0 µM Cu II) growth condition for the 16, 22, or 32 hr dataset divided by the standard deviation for the 16, 22, or 32 hr dataset.
There are 783 samples in our dataset, divided into 29 different experimental series (see Table 1).
BI was implemented with the program Mr.Bayes v3.1.2 [ 59] under a partitioned model (dataset divided into genes), and considering the model of nucleotide substitution estimated with jModelTest.
Patients were sorted in descending order by these predicted probabilities, and the resultant dataset divided into terciles (3 equinumerous, non-overlapping, exhaustive subsets).
The apparent misclassification error rate, which is the number of misclassified observations in the training dataset divided by the total number of samples in the training dataset, tends to under-estimate the true misclassification error rate [ 32].
The specificity of each approach for CXCR4 prediction was calculated as the number of predicted R5 viruses in the CCR5-using dataset divided by the total number of sequences in the CCR5-using dataset.
The sensitivity of each approach for CXCR4 prediction was calculated as the number of predicted X4 viruses in the CXCR4-using dataset divided by the total number of sequences in the CXCR4-using dataset.
This file gives the position of the QTL detected with the dataset divided by diet, with the top SNP corresponding to the SNP with the highest effect in the QTL region.
