Software note: using probe secondary structure information to enhance Affymetrix GeneChip background estimates

Abstract

High-density short oligonucleotide microarrays are a primary research tool for assessing global gene expression. Background noise on microarrays comprises a significant portion of the measured raw data. A number of statistical techniques have been developed to correct for this background noise. Here, we demonstrate that probe minimum folding energy and structure can be used to enhance a previously existing model for background noise correction. We estimate that probe secondary structure accounts for up to 3% of all variation on Affymetrix microarrays.

Fig. 1

R² distribution for the simple linear models (Eq.s 2-4) for all the U133 Latin Square chips. The null hypothesis that the B1 parameter is equal to zero is rejected with high confidence (P < 10-4) for all the models.

Fig. 2

Folded probe showing its sequence, minimum folding energy (ΔG_ss) and minimum energy structure. The longest free string of bases of the folded probe (S_L) is shown in gray box; bases involved in hydrogen bonding are shown in green ovals.

Fig. 3

A model for the relationship between ΔG_ss and θ for the bases involved in secondary structure formation.

Fig. 4

Effects of changing the values of ΔG_ss-cutoff and tb on the performance of the position-dependent secondary-structure attenuated affinity model using the human genome U133 Latin Square Experiment 2 Replicate 1 PM probes.

Fig. 5

Effects of changing tb or ΔG_ss-cutoff on the performance of the PSAA model. Effects of changing (A) the values of tb while holding ΔG_ss-cutoff = -3.6 or (B) the values of ΔG_ss-cutoff while holding tb = 0.35 on the performance of PSAA: the position-dependent secondary-structure attenuated affinity model (Eq. 6). Data shown are for the human genome U133 Latin Square Experiment 2 Replicate 1 PM probes. NM: Naef and Magnasco (Naef and Magnasco, 2003) model (Eq. 5). The suffixes (-SH) and (-RD) indicates the R² after generating the minimum folding energy (ΔG_ss) and the minimum energy structure from shuffled and random sequences, respectively (see section 3.3 for explanation).