Genes were then sorted according to values of the topological indices.
To perform GSEA fold changes were generated to compare two samples, which were then sorted according to values from highest to lowest.
The prediction results from the same method were further sorted according to E-values (sequence alignment-based prediction, and domain-based prediction) or prediction scores (machine-learning prediction).
After having verified that the fitted models did not show overdispersion, these were sorted according to increasing values of the second-order bias correction Akaike information criterion (AICc), as appropriate for models in which the ratio of number of observations n to number of parameters K is n/ K ≤ 40.
The compound list at the left-hand side shows the logD value for each compound and the list is sorted according to the logD value.
The tissue-specific gene targets were sorted according to the average value of classifier outputs from ten predictions, and a higher value might indicate a higher probability of being expressed predominantly in a particular tissue.
The features are sorted according to their IG value.
Then, the images are sorted according to the similarity values which make up a ranking.
When csNAPs of E. coli in the log phase are sorted according to the emPAI values that roughly represent the amount of proteins in the mixture [48], various DNA and/or RNA binding proteins, such as Hu (hupA, hupB) and StpA (stpA), are included in the top 10 E. coli csNAPs (Figure 6A).
Listed in Table 1 and 2 are the top positions sorted according to their U-values for RSS (stringent cut-off of 15) and Z-scores for CSS (stringent cut-off of 2), respectively.
An algorithm for computing the conformity is based on the following Lemma 11 If A denotes a symmetric real matrix with eigenvalues d1,..., d n sorted according to their absolute values, i.e., | d1| ≥ | d2| ≥... ≥ | d n |, and the corresponding orthonormal eigenvectors are denoted by u 1,..., u n, then ‖ A − v v τ ‖ F 2 is minimized at v* = | d 1 | u 1.
