The E-MAP dataset was already heavily pre-processed and custom-scored, as has been noted also in previous works [26], and therefore it was not expected that the matrix decomposition could provide marked improvements in this data.
A second map dataset was contributed by Prof. Juan F. Medrano and after size filtering (≥300 kb) it added another 1,064,919 Rmaps, bringing the total raw dataset up to 2,973,315 (439 X coverage, before alignment).
We run two additional PC Analyses in which Hap Map datasets were included (European datasets only in a first instance and then Hap Map datasets from European, African and Asian ancestry).
Two E-MAP datasets were used here, in which the mapping approach was applied to the genes involved either in the yeast early secretory pathway (Secretory-Map) [ 53] or in various aspects of chromosome biology (Chromosome-Map) [ 54].
For each position to which a probe pair was mapped, a dataset was generated consisting of all pairs mapping within a window of ± 30 bp, which defines the number of bases extending from the position being analyzed so that every probe in the 61-bp region was included in the signal and p-value analysis.
We used SNP genotypes to characterize the genomes of 288 NMRI mice to assess their suitability as a mapping population (the final dataset is available at http://cgd.jax.org/datasets/datasets.shtml).shtml
In a preprocessing step, the dataset is mapped into a virtual three-dimensional space.
When performing visual validation with CheS-Mapper the dataset is mapped into 3D space based on the same features as employed by the (Q SAR modeling.
The top-scored map for the combined dataset is the IFN-gamma signaling pathway (Table 1).
Before feeding the classifier with the dataset, the dataset is mapped using Linear Discrimination Analysis (LDA) and then divided into training and testing sets.
