|
|
Blood, Vol. 109, Issue 5, 2066-2077, March 1, 2007
Gene expression patterns in blood leukocytes discriminate patients with acute infections
Blood Ramilo et al.
109: 2066
Supplemental materials for: Ramilo et al, Vol 109, Issue 5, 2066-2077
Files in this Data Supplement:
- Document 1. Supplemental methods: real-time RT-PCR analysis (PDF, 62.1 KB)
- Document 2. Supplemental methods: K-Nearest Neighbors (kNN) (PDF, 11.2 KB)
- Figure S1. Validation of differentially expressed genes by real-time RT-PCR (JPG, 336 KB)
-
(A) Levels of expression of 9 genes were measured by real-time reverse transcription–polymerase chain reaction (RT-PCR) in samples obtained from patients with S aureus (Sa) or E coli (Ec) infections (fold change in gene expression over healthy controls, log transformed except for RGS2, FCAR, and ALOX). Each plot represents one sample; lines represent median expression. (B) Correlation between expression values obtained by real-time RT-PCR analysis (abscissa) and microarray analysis (ordinate, normalized to the expression in the sample from the same healthy control to which real-time RT-PCR data were normalized; log scale). See Table S5 for details.
- Figure S2. Discriminating patients with bacterial infections (JPG, 575 KB)
-
(A) Hierarchical clustering of 242 genes obtained from Mann-Whitney rank test comparison (P < .01) between groups of patients with E coli infections (11 samples) and S pneumoniae infections (11 samples). Transformed expression levels are indicated by color scale, with red representing relative high expression and blue indicating relative low expression compared to the median expression for each gene across all donors. Genes are listed in Table S7. (B) A supervised learning algorithm was used to identify genes representing the highest capacity to discriminate the two classes. Leave-one-out cross-validation of the training set with 45 predictor genes classified the samples with 85% (20 of 22) accuracy. Classifier genes are listed in Table S8. (C) Unsupervised hierarchical clustering of 127 genes obtained from Mann-Whitney rank test comparison (P < .01) between groups of patients with S aureus infection (12 samples) or S pneumoniae infection (11 samples). Transformed expression levels are indicated by color scale, with red representing relative high expression and blue indicating relative low expression compared to the median expression for each gene across all donors. Genes are listed in Table S9. (D) A supervised learning algorithm was used to identify genes presenting the highest capacity to discriminate the 2 classes. Leave-one-out cross-validation of the training set with 30 genes classified the samples with 83% (19 of 23) accuracy. Classifier genes are listed in Table S10.
- Table S1. Comparison of clinical parameters between groups of patients included in the training sets (Mann-Whitney test) (XLS, 41 KB).
- Table S2. Influenza A versus bacterial infection; class comparison results (XLS, 225 KB)
-
Comparison of gene expression levels (normalized values) between patients with acute influenza A infection and patients with acute bacterial infection (E coli or S pneumoniae); P < .01, Mann-Whitney test, 854 genes. Expression patterns are represented in Figure 1A.
- Table S3. Influenza A versus bacterial infection; classifier genes (XLS, 19 KB)
-
List of 35 classifier genes discriminating samples obtained from patients.
- Table S4. S aureus versus E coli; class comparison results (XLS, 66 KB)
-
Comparison of gene expression levels (normalized values) between patients with acute S aureus infection and patients with acute E coli infection; P < .01, Mann-Whitney test, 211 genes. Expression patterns are represented in Figure 3A.
|
|
