Machine learning-based microarray analyses indicate low-expression genes might collectively influence PAH disease.

The enhanced modeling performance, particularly in the top 200 genes range, caused by SIRRFE were extremely significant under All Genes ( Fig 3A ), All>12 as identified by the IQR method ( Fig 3B ), ALOGA<128 ( Fig 3H ), and ALOGA<256 ( Fig 3H ) filtering methods; and moderately significant under ALOGA>128 ( Fig 3E ) and ALOGA>256 ( Fig 3F ) methods.

also in 2,178 other papers