The prediction accuracy of the 54 validation samples, using different gene selection procedures, is summarized in Figure 2. The results showed that fold change and penalized t-statistic based methods for feature selection outperformed the t-statistic-based procedure.
A scalable automated non-gel based method for size selection of DNA molecules have been designed to replace the laborious and time consuming agarose gel separation step that dramatically increases sample throughput for massive sequencing, and are suitable for all similar DNA processing protocols demanding high throughput and a controllable size interval.
It is suggested that in theory amino acid-based methods for selection tests may better contain divergent sequences and dS saturation than codon-based methods [ 45].
Also, what types of kernels were used and why? 4) Section entitled "Causal graph-based methods for selection of SNPs", first paragraph.
5) We added the description of the TIE* algorithm to the Additional File 2 and also revised the algorithm description in the manuscript, please see first paragraph of the sub-section "Causal graph-based methods for selection of SNPs".
2) The discussion of false positives and false negatives in the output of the employed causal discovery algorithm is provided in the last paragraph of sub-section "Causal graph-based methods for selection of SNPs".
We set the maximum conditioning set size parameter max-k = 3 primarily because of the sample size limitations as described in the fourth paragraph of sub-section "Causal graph-based methods for selection of SNPs" in the revised manuscript.
2. A) We added the description of the TIE* algorithm to the Additional File 2 and also revised the algorithm description in the manuscript, please see first paragraph of the sub-section "Causal graph-based methods for selection of SNPs".
Traditional case-mix adjustments via multivariate regression based methods cannot completely account for the selection bias.
For without feature selection case, using only frequent itemsets based methods performed worse compared to base case in terms of f1 score and accuracy.
