Blood samples were drawn around mid-day in all cases to reduce the potential contribution of the variation in gene expression during different times of day. Normal healthy twin pairs were recruited for the study. Three pair of female twins belonged to age group 20-23 years and two pair of male twins were of 25 and 37 years of age. Three more normal individuals including two females and one male were recruited. Their ages were 23, 34 and 37 years respectively. Informed consent was obtained from all. About 20 ml of blood was drawn by vein puncture and immediately processed for nucleic acids isolation. A 3/4th of the isolated blood was used for total RNA isolation and the rest was used for isolating genomic DNA. Twelve highly polymorphic microsatellite markers, located on 8 different chromosomes (Perkin Elmer Linkage panel set version 2, PE Applied Biosystems, Foster City, CA) were used for haplotyping of genomic DNA from twins to assess their monozygosity (Sharma et al., 2005).

Confirmation of zygosity of twins
All the five twin pairs had identical alleles for the 12 repeat markers. Since the probability of monozygosity is greater than 99.9% when more than five highly polymorphic markers have identical size distribution within a twin pair (Amann et al., 2001), our data confirms that the five pairs of twins are monozygotic.

Isolation of total RNA and genomic DNA from blood leukocytes
Total RNA was isolated from blood leukocytes after the red blood cells were lysed in 1X RBC lysis buffer (150mM NH4Cl, 10 mM NaHCO3 1mM EDTA prepared in DEPC treated water). The blood leukocytes were recovered by centrifugation at 250g and total RNA isolated using EZ-RNA isolation kit (Biological Industries, Israel). The Genomic DNA was isolated using the salting out procedure (Miller et al., 1988).

Data Submission
The raw data from the GeneChip experiments has been submitted to Gene expression Omnibus (GEO; www.ncbi.nlm.nih.gov/geo) under the GEO series number GSE928.

Ethical Clearance
Ethical clearance for the project was obtained from Human Ethics Committee of the Institute. Informed consent was obtained from all the subjects.

Labeled cRNA preparation
The amount of RNA taken from each sample was equalized based on absorbance at 260nm. Double stranded cDNA was synthesized from 8µg of total RNA by reverse transcription using T7-(dT)24 primer and the Superscript Choice cDNA synthesis system (Invitrogen, USA). In vitro transcription of the cDNA was carried out using Enzo Bioarray High Yield RNA transcript labeling kit (Affymetrix Inc., USA) to prepare biotin labeled cRNA. The labeled cRNA was cleaned using RNAeasy columns (Qiagen, USA). The labeled target was fragmented, and a hybridization cocktail prepared including fragmented cRNA, probe array controls, BSA and Herring sperm DNA.

Gene Chip processing
Gene chips were processed (HG U95Av2 arrays, Affymetrix Inc., USA) under same set of experimental conditions. First, labeled products were hybridized with the Affymetrix GeneChip Test3 arrays. If the results were judged satisfactory, hybridization was subsequently carried out with the HG U95Av2 arrays as per manufacturer's instructions. Arrays were hybridized at 45°C for 16 hours. After hybridization, arrays were washed using an automated Gene Chip Fluidics Station 400. After washing, the array was stained with streptavidin-phycoerythrin and scanned using HP Gene Array Scanner. Data analysis was carried out using Affymetrix Microarray Suite Software (MAS 5). All Gene Chip experiments were performed at the Weizmann Institute of Science, Rehovot, Israel.

In this release, the database contains gene expression data of blood leukocytes from 13 normal human individuals (five pairs of monozygotic twins and 3 unrelated individuals) measured using HG-U95A oligonucleotide microarrays consisting probes for ~10,000 genes.
The data was analyzed using GeneChip software(Affymetrix) for expression analysis (MAS 5). Global scaling was carried out to reliably compare the data from multiple arrays. In pairwise comparisons, probe sets with Detection call of ‘Present’ (P) and Detection p-value range: 0-0.04 were considered. The differentially expressed genes were identified by selecting the probe sets with ‘Change’ call of either ‘I’ or ‘D’ and with a signal log ratio of > 1.585. We used signal log ratio value of 1.585 (absolute value) as cutoff because, none of the genes from ~10,000 genes had a signal log ratio of greater than 1.585 in duplicate experiments using the same RNA sample.
The mean expression of each gene was computed from the ‘signal’ values across all 13 arrays, and taking the log10 transform of the signal values. Multiple probe sets if present, for a gene were included if they had ‘P’ call and their log10 transformed signal values were averaged. Probe sets without ‘P’ calls were not considered. The Coefficient of variation (CV) was computed as SD/Mean where SD is the standard deviation of the log10 transformed signal values across the arrays. Annotation information on the functional roles of the differentially expressed genes was mainly obtained from NetAffx from Affymetrix (www.affymetrix.com), GeneCards (http://www.genecards.org), and Entrez Gene at the NCBI Web site (www.ncbi.nlm.nih.gov/). Gene symbols and their descriptions corresponding to each probeset were obtained from annotation file of HG U95Av2 (version 23rd June 2004) from Affymetrix web site (www.affymetrix.com).

Annotation information on the functional roles of the differentially expressed genes was obtained from the GeneCards, Entrez gene and NetAffx from Affymetrix Web site.

Expression of genes in different human tissues was obtained from UniGene database (Build 160). The Expressed Sequence Tags (ESTs) libraries of blood were retrieved and the genes were classified into highly expressing (H), moderately expressing (M) and weakly expressing (W) using the criteria described by Bortoluzzi et al. 2000.

Information on the housekeeping genes was obtained from GNF (Gene Expression Atlas Database) data(http://expression.gnf.org). Eisenberg and Levanon (2003) have identified a set of 575 housekeeping genes from the publicly available microarray database, GNF. The gene expression profiles were obtained from a set of 46 tissues, organs and cell lines by using Human U95 A Gene Chip (Affymetrix Inc, USA). The list of housekeeping genes was obtained from the site (http://www.compugen.co.il/supp_info/Housekeeping_genes.html). There were 559 genes common between the housekeeping gene list and the genes with known symbols present on HG U95Av2 arrays. Probe set for the rest of the genes was not found on HG U95Av2 array. Of these 559 genes, a total of 542 genes were found expressed in blood in our experiments.Related information on the annotation and function of these genes was obtained from NCBI Entrez Gene (http://www.ncbi.nlm.nih.gov/) and GeneCards (http://www.genecards.org) databases. Information on the roles of gene in biochemical pathways as outlined in KEGG (Kyoto encyclopedia of genes and genomes) (http://www.genome.ad.jp/kegg/kegg2.html) and GenMAPP (http://www.genmapp.org) databases was obtained from Affymetrix (http://www.affymetrix.com/index.affx).