Sentence examples for predictions we selected from inspiring English sources

From the remaining 119 DTIs, which are from the viewpoint of our method new predictions, we selected 15 (Table 3) with the highest prediction score and evaluated them manually.

Relying on in silico computer modeling predictions, we selected five different linker regions from the RCSB protein database that connect oligomerization domains, and then further studied the self-assembly and stability of in vitro produced nanoparticles through biophysical characterization of formed particles.

To prove the reliability of our predictions, we selected 5 of these 246 molecules (ZINC00058225, ZINC01669260 and ZINC16946275 from Cluster 1; ZINC03683886 from Cluster 2; and ZINC03871389 from Cluster 3; see Figure 4) and tested their effects on the hIKK-2 activity using an in vitro assay (see Figure 5).

To confirm these predictions we selected 14 genes for detailed analyses by bisulfite sequencing.

For both sets of predictions, we selected Pareto-optimal predictions and ran MITIE on the corresponding graphs.

To validate our bioinformatics-based predictions, we selected 30 genes for expression analysis using specific cancer cell lines.

Using a low cut-off algorithm-based prediction, we selected 256 potential HLA-A3 ligands from four melanoma-associated proteins.

To provide a real world example of distributed metabolic pathway prediction we selected a symbiotic system with known nutritional provisioning requirements.

To validate the results obtained by computational prediction, we selected two probes for the two microRNA families hsa-mir-1233 hsa-mir-1233 hsa-mir-1233beled them with TandA (red) and FITC (green), respectively (Additional file 11).

For ER-status prediction, we selected QDA with FS1 (conditional validation AUC = 0.939); for pCR prediction including both the ER-positive and -negative cancers, we selected LREG with FS5 (conditional validation AUC = 0.805); and for pCR in ER-negative cancers, we selected LREG with FS4 (conditional validation AUC = 0.627).

Development of the prediction model: We selected predictive descriptors using feature selection algorithms (provided by RapidMiner 5.1.13), which returned 113 descriptors as presumably predictive.