Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

ReliefSeq: A Gene-Wise Adaptive-K Nearest-Neighbor Feature Selection Tool for Finding Gene-Gene Interactions and Main Effects in mRNA-Seq Gene Expression Data

Figure 3

Illustration of the effect of the number of nearest neighbors (k) on the Relief-F score for two interacting transcripts (red circles and green triangles) from one of the simulated data sets.

A lower k is optimal for interacting transcripts. For larger k, Relief-F converges to a univariate attribute estimator (myopic) and the scores of the interacting transcripts tend toward zero because they have no main effect. For comparison, the average scores and 5th and 95th percentiles for 100 random null transcripts (blue diamonds) in the data set are plotted versus k. The Relief-F scores for noise genes are low and relatively unaffected by the value of k.

Figure 3