Figure 1

Flow diagram of the S-Lowess procedure. The S-Lowess procedure consists of two phases: 1) determine or upload likely conserved genes (LCG) and 2) Normalize a microarray dataset with the LCGs. In case that for phase 1 de novo prediction of LCGs is selected, the user has to upload microarray feature sequences and select (multiple) genomes (in this study 3 Streptococcus genomes). The optimal parameters for selection of LCGs from a sequence comparison using BLAT of array features versus multiple reporter genomes are difficult to predict. Therefore, selection of a LCG set is facilitated by cycling through a maximum of 2 parameters. These parameters are (a combination of two): (i) alignment length cutoff, (ii) E-value cutoff, (iii) percentage nucleotide identity cutoff, (iv) maximum number of hits within the same genome (to account for paralogous genes or duplicated genome fragments), (v) minimum number of hits across genomes (to account for gene conservation in multiple genome sequences). Those array feature sequences meeting the criteria (here in at least 2 out of three genomes a significant BLAT hit; one hit over at least 100 bp with at least 80% nucleotide identity) are marked as LCG and added to the conserved array feature list. In phase 2, the LCGs are used to normalize an uploaded aCGH microarray dataset. The result of phase 2 is a normalized dataset and diagnostic plots.