I/NI-calls for the exclusion of non-informative genes: a highly effective feature filtering tool for microarray data

Willem Talloen, Djork-Arné Clevert, Sepp Hochreiter, Dhammika Amaratunga, Luc Bijnens, Stefan Kass and Hinrich W.H. Göhlmann
Bioinformatics 2007, 23(21): 2897-2902, doi:10.1093/bioinformatics/btm478 (online)

DNA microarray technology typically generates many measurements of which only a relatively small subset is informative for the interpretation of the experiment. To avoid false positive re-sults, it is therefore critical to select the informative genes from the large noisy data before the actual analysis. Most currently available filtering techniques are supervised and therefore suffer from a po-tential risk of overfitting. The unsupervised filtering techniques, on the other hand, are either not very efficient or too stringent as they may mix up signal with noise. We propose to use the multiple probes measuring the same target mRNA as repeated measures to quantify the signal-to-noise ratio of that specific probe set. A Bayes-ian factor analysis with specifically chosen prior settings, that mod-els this probe level information, is providing an objective feature filtering technique, named I/NI calls.

Based on 30 real-life data sets (including various human, rat, mice and Arabidopsis studies) and a spiked-in data set, it is shown that I/NI calls is highly effective, with exclusion rates ranging from 70 to 99%. Consequently, it offers a critical solution to the curse of high-dimensionality in the analysis of microarray data.

This filtering approach is publicly available as a function implemented in the R package FARMS

Informative/Non-informative calls (I/NI calls), Absent/Present calls (A/P calls), FARMS (Factor Analysis for Robust Microarray Summarization)