Methodology article
A comparison on effects of normalisations in the detection of differentially expressed genes
-
* Corresponding author: Chiara Romualdi chiara.romualdi@unipd.it
1 Department of Statistical Sciences, University of Padova, via C. Battisti 241, 35121 Padova, Italy
2 CRIBI biotechnology Center, University of Padova, via U. Bassi 58/B, 35121 Padova, Italy
BMC Bioinformatics 2009, 10:61 doi:10.1186/1471-2105-10-61
Published: 13 February 2009Additional files
Additional file 1:
Figure S1. Examples of typical MA plots obtained with LNN and GG models without systematic bias (panel A and B, respectively), with LNN and GG models with systematic bias (panel C and D, respectively) and with Albers'™ model with negative values (panel E) and with negative values replaced by positive constant (panel F). Red points represent differentially simulated expressed genes.
Format: PDF Size: 908KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 2:
Figure S2. Specificity and sensitivity curves after quantile, qspline, quantile enhanced and qspline enhanced for Albers' model with increasing percentage of background level with respect to expression level with and without replacing negative values.
Format: PDF Size: 83KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 3:
Figure S3. LNN and GG models without non-linear bias. Specificity and sensitivity (panel A and B) and average overlapping rates of top ranking gene lists detected as differentially expressed between lowess and the other normalisations.
Format: PDF Size: 22KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 4:
Figure S4. Specificity and sensitivity curves and overlapping rates of top ranking gene lists for Albers' model with increasing percentage of background level with respect to expression level with and without replacing negative values, using the moderated t-test EBayes.
Format: PDF Size: 250KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 5:
Table S5. Area Under the Curve (AUC) of specificity and sensitivity of the moderated t-test EBayes after the normalisations, for Albers' model with increasing percentage of background level with and without replacing negative values. For each simulated scenario, ranking of the normalisations according to the AUC is reported.
Format: PDF Size: 18KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 6:
Figure S5. Overlapping rates of top ranking gene lists detected as differentially expressed between lowess and the others normalisations with 95% empirical confidence interval. Panel A, C, E: results obtained from data generated by Albers' model with 10%, 50% and 150%, respectively, background levels without negative values replacement; panel B, D, F: results obtained from Albers' model with 10%, 50% and 150%, respectively, background levels with negative values replacement.
Format: PDF Size: 54KB Download file
This file can be viewed with: Adobe Acrobat Reader
Additional file 7:
Table S2. Parameters setting used in the Albers' simulation model.
Format: PDF Size: 106KB Download file
This file can be viewed with: Adobe Acrobat Reader