Mickael Guedj

A comprehensive evaluation of normalization methods for Illumina high-throughput RNA sequencing data analysis

During the last 3 years, a number of approaches for the normalization of RNA sequencing data have emerged in the literature, differing both in the type of bias adjustment and in the statistical strategy adopted. However, as data continue to accumulate, there has been no clear consensus on the appropriate normalization method to be used or the impact of a chosen method on the downstream analysis. In this work, we focus on a comprehensive comparison of seven recently proposed normalization methods for the differential analysis of RNA-seq data, with an emphasis on the use of varied real and simulated datasets involving different species and experimental designs to represent data characteristics commonly observed in practice. Based on this comparison study, we propose practical recommendations on the appropriate normalization method to be used and its impact on the differential analysis of RNA-seq data.

A refined molecular taxonomy of breast cancer

The current histoclinical breast cancer classification is simple but imprecise. Several molecular classifications of breast cancers based on expression profiling have been proposed as alternatives. However, their reliability and clinical utility have been repeatedly questioned, notably because most of them were derived from relatively small initial patient populations. We analyzed the transcriptomes of 537 breast tumors using three unsupervised classification methods. A core subset of 355 tumors was assigned to six clusters by all three methods. These six subgroups overlapped with previously defined molecular classes of breast cancer, but also showed important differences, notably the absence of an ERBB2 subgroup and the division of the large luminal ER+ group into four subgroups, two of them being highly proliferative. Of the six subgroups, four were ER+/PR+/AR+, one was ER−/PR−/AR+ and one was triple negative (AR−/ER−/PR−). ERBB2-amplified tumors were split between the ER−/PR−/AR+ subgroup and the highly proliferative ER+ LumC subgroup. Importantly, each of these six molecular subgroups showed specific copy-number alterations. Gene expression changes were correlated to specific signaling pathways. Each of these six subgroups showed very significant differences in tumor grade, metastatic sites, relapse-free survival or response to chemotherapy. All these findings were validated on large external datasets including more than 3000 tumors. Our data thus indicate that these six molecular subgroups represent well-defined clinico-biological entities of breast cancer. Their identification should facilitate the detection of novel prognostic factors or therapeutical targets in breast cancer.

Association between the IRF5 rs2004640 functional polymorphism and systemic sclerosis: A new perspective for pulmonary fibrosis

OBJECTIVE There is now growing evidence that connective tissue diseases, including systemic sclerosis (SSc), share a common genetic background. Microarray studies support a pivotal role of type I interferon (IFN) in the pathophysiology of connective tissue diseases. Interferon regulatory factors coordinate the expression of type I IFNs, and the IRF5 gene has been identified as a susceptibility gene of systemic lupus and Sjögrens syndrome. The aim of this study was to determine whether the IRF5 rs2004640 single-nucleotide polymorphism is associated with SSc. METHODS The IRF5 rs2004640 (GT) functional polymorphism was genotyped in 1,641 subjects of French European Caucasian origin: a discovery set comprising 427 patients with SSc and 380 control subjects and a replication set comprising 454 patients with SSc and 380 control subjects. RESULTS In both the discovery set and the replication set, the TT genotype was significantly more common in patients with SSc than in control subjects, with an odds ratio (OR) for the combined populations of 1.58 (95% confidence interval [95% CI] 1.18-2.11 [P for trend 0.002]). Analyses of the whole SSc population showed a significant association between homozygosity for the T allele and the presence of antinuclear antibodies (corrected P [Pcorr]=0.04, OR 1.59, 95% CI 1.16-2.17) and fibrosing alveolitis (Pcorr=0.001, OR 2.07, 95% CI 1.38-3.11). In a multivariate analysis model including the diffuse cutaneous subtype of SSc and positivity for anti-topoisomerase I antibodies, the IRF5 rs2004640 TT genotype remained associated with fibrosing alveolitis (P=0.029, OR 1.92, 95% CI 1.07-3.44). CONCLUSION The IRF5 rs2004640 GT substitution is associated with susceptibility to SSc. These data provide new insight into the pathogenesis of SSc, including clues to the mechanisms leading to fibrosing alveolitis.

STAT4 is a genetic risk factor for systemic sclerosis having additive effects with IRF5 on disease susceptibility and related pulmonary fibrosis

OBJECTIVE Systemic sclerosis (SSc) belongs to the group of connective tissue disorders (CTDs), among which are several disorders characterized by a type I interferon (IFN) signature. The recent identification of an association between IRF5 and SSc further highlights a key role for IFN. STAT4, which encodes STAT-4, contributes to IFN signaling, and its genetic variants were found to be associated with CTDs. The aim of this study was to determine whether the STAT4 rs7574865 single-nucleotide polymorphism is associated with SSc, and whether it interacts with IRF5. METHODS Both the STAT4 rs7574865 and IRF5 rs2004640 polymorphisms were genotyped in 1,855 individuals of French Caucasian origin comprising a discovery set of 440 patients with SSc and 485 control subjects and a replication set of 445 patients with SSc and an additional 485 control subjects. RESULTS STAT4 rs7574865 was shown to be associated with SSc (P=0.001, odds ratio [OR] 1.29, 95% confidence interval [95% CI] 1.11-1.51). This association was not restricted to a particular phenotype. An additive effect of the STAT4 rs7574865 T allele and the IRF5 rs2004640 T allele was observed, resulting in a multiple increased 1.28-fold risk of SSc. The OR for SSc was 2.72 (95% CI 1.86-3.99) for combinations of genotypes with >or=3 risk alleles. An additive effect was also detected for fibrosing alveolitis: carriage of at least 3 risk alleles appeared to be an independent risk factor (P=2.2x10(-4), OR 1.97, 95% CI 1.28-3.04). CONCLUSION Our results establish STAT4 rs7574865 as a new SSc genetic susceptibility factor. STAT4 and IRF5 act with additive effects in terms of susceptibility to both SSc and SSc-related fibrosing alveolitis.

The PTPN22 620W allele confers susceptibility to systemic sclerosis: Findings of a large case–control study of European Caucasians and a meta‐analysis

OBJECTIVE To determine whether genetic variants of the PTPN22 gene, including the R620W (1858C>T) missense single-nucleotide polymorphism (SNP), are associated with systemic sclerosis (SSc). Since PTPN22 is involved in multiple autoimmune diseases, we also examined the occurrence of a concomitant autoimmune disease. We then conducted a meta-analysis of the most recent studies of SSc in order to verify the association or lack of association between the PTPN22 1858C>T variant and SSc. METHODS Seven PTPN22 SNPs were analyzed in a French Caucasian cohort of 659 SSc patients and 504 healthy controls. All SSc patient sera were tested for the presence of autoantibodies against topoisomerase I (anti-topo I) and for anticentromere antibodies (ACAs). RESULTS The co-occurrence of an autoimmune disease was observed in 22% of the 416 SSc patients who were exhaustively screened. In 33 of the 416 patients (8%), the concomitant autoimmune disease was known to be associated with PTPN22 1858T; these patients were excluded prior to analysis. No association was detected for any of the SNPs tested. PTPN22 haplotype analysis identified a strong association between SSc and the presence of a risk haplotype carrying the 1858T allele (P = 1.52 x 10(-7)) and a protective haplotype carrying the 1858C allele (P = 2.20 x 10(-16)) in our French Caucasian population. The meta-analysis provided evidence that the PTPN22 1858T allele is involved in the genetic susceptibility to SSc in Caucasian (P = 8.39 x 10(-3), OR 1.08 [95% CI 1.02-1.15]) and mixed (P = 3.11 x 10(-3), OR 1.09 [95% CI 1.04-1.16]) populations, particularly in the anti-topo I-positive subset. CONCLUSION Our results indicate that PTPN22, a shared genetic factor of multiple autoimmune diseases, also contributes to the genetic background of SSc.

Association of the TNFAIP3 rs5029939 variant with systemic sclerosis in the European Caucasian population

Background TNFAIP3 encodes the ubiquitin-modifying enzyme, a key regulator of inflammatory signalling pathways. Convincing associations between TNFAIP3 variants and autoimmune diseases have been reported. Objective To investigate the association of TNFAIP3 polymorphisms with systemic sclerosis (SSc). Methods Three single nucleotide polymorphisms (SNPs) in a set of 1018 patients with SSc and 1012 controls of French Caucasian origin were genotyped. Two intergenic SNPs, rs10499194 and rs6920220, and one located in TNFAIP3 intron 2, rs5029939, were selected. The TNFAIP3 rs5029939 found to be associated with SSc in this first set was then genotyped in a second set of 465 patients with SSc and 182 controls from Germany and 184 patients with SSc and 124 controls from Italy. Pooled odd ratios were calculated by Mantel–Haenszel meta-analysis. Results The rs5029939 G allele was found to be significantly associated with SSc susceptibility (pooled OR=2.08 (95% CI 1.59 to 2.72); p=1.16×10−7), whereas the rs10499194 and rs6920220 variants displayed no association. Only one of the predicted haplotypes investigated in the French sample was significantly associated with SSc (p=8.91×10−8), and this haplotype was discriminating only in the presence of the rs5029939 risk allele, suggesting that this SNP tags the association signal. The strongest associations of rs5029939 with subphenotypes, having large magnitudes for complex genetic disorders, were observed for diffuse cutaneous SSc (pooled OR=2.71 (1.94 to 3.79), p=5.2×10−9), fibrosing alveolitis (pooled OR=2.26 (1.61 to 3.17), p=2.5×10−6) and pulmonary arterial hypertension (pooled OR=3.11 (1.86 to 5.17), p=1.3×10−5). Conclusion These results suggest that TNFAIP3 is a genetic susceptibility factor for SSc.

Should We Abandon the t-Test in the Analysis of Gene Expression Microarray Data: A Comparison of Variance Modeling Strategies

High-throughput post-genomic studies are now routinely and promisingly investigated in biological and biomedical research. The main statistical approach to select genes differentially expressed between two groups is to apply a t-test, which is subject of criticism in the literature. Numerous alternatives have been developed based on different and innovative variance modeling strategies. However, a critical issue is that selecting a different test usually leads to a different gene list. In this context and given the current tendency to apply the t-test, identifying the most efficient approach in practice remains crucial. To provide elements to answer, we conduct a comparison of eight tests representative of variance modeling strategies in gene expression data: Welchs t-test, ANOVA [1], Wilcoxons test, SAM [2], RVM [3], limma [4], VarMixt [5] and SMVar [6]. Our comparison process relies on four steps (gene list analysis, simulations, spike-in data and re-sampling) to formulate comprehensive and robust conclusions about test performance, in terms of statistical power, false-positive rate, execution time and ease of use. Our results raise concerns about the ability of some methods to control the expected number of false positives at a desirable level. Besides, two tests (limma and VarMixt) show significant improvement compared to the t-test, in particular to deal with small sample sizes. In addition limma presents several practical advantages, so we advocate its application to analyze gene expression data.

BANK1 is a genetic risk factor for diffuse cutaneous systemic sclerosis and has additive effects with IRF5 and STAT4

OBJECTIVE To determine whether the functional BANK1 variants rs3733197 and rs10516487 are associated with systemic sclerosis (SSc) in 2 European Caucasian populations and to investigate the putative gene-gene interactions between BANK1 and IRF5 as well as STAT4. METHODS BANK1 single-nucleotide polymorphisms were genotyped in a total population of 2,432 individuals. The French cohort consisted of 874 SSc patients and 955 controls (previously genotyped for both IRF5 rs2004640 and STAT4 rs7574865). The German cohort consisted of 421 SSc patients and 182 controls. RESULTS The BANK1 variants were found to be associated with diffuse cutaneous SSc (dcSSc) in both cohorts, providing an odds ratio (OR) of 0.77 for the rs10516487 T rare allele in the combined populations of dcSSc patients as compared with the combined populations of controls (95% confidence interval [95% CI] 0.64-0.93) and an OR of 0.73 (95% CI 0.61-0.87) for the rs3733197 A rare allele. BANK1 haplotype analysis found the A-T haplotype to be protective in dcSSc patients (OR 0.70 [95% CI 0.57-0.86], P = 3.39 x 10(-4)) and the G-C haplotype to be a risk factor (OR 1.25 [95% CI 1.06-1.47], P = 0.008). Significant differences were also observed when the limited cutaneous subset of SSc was compared with the dcSSc subset, both for the rare alleles and for the haplotypes. The BANK1, IRF5, and STAT4 risk alleles displayed a multiplicatively increased risk of dcSSc of 1.43-fold. CONCLUSION Our results establish BANK1 as a new SSc genetic susceptibility factor and show that BANK1, IRF5, and STAT4 act with additive effects.

NLRP1 influences the systemic sclerosis phenotype: a new clue for the contribution of innate immunity in systemic sclerosis-related fibrosing alveolitis pathogenesis

Background Recent evidence has highlighted a potential role of interleukin 1β (IL-1β) in systemic sclerosis (SSc). NLRP1 provides a scaffold for the assembly of the inflammasome that promotes the processing and maturation of pro-IL-1β. In addition, NLRP1 variants were found to confer susceptibility to autoimmune disorders. Objective To study a possible association of the NLRP1 rs6502867, rs2670660 and rs8182352, rs12150220 and rs4790797 with SSc in the European Caucasian population. Methods NLRP1 single nucleotide polymorphisms were genotyped in 3227 individuals comprising a discovery set (870 SSc patients and 962 controls) and a replication set including individuals from Germany (532 SSc patients and 324 controls) and Italy (527 SSc patients and 301 controls), all individuals being of European Caucasian origin. Results Conditional analyses revealed a significant association for the NLRP1 rs8182352 variant with both anti-topoisomerase-positive and SSc-related fibrosing alveolitis (FA) subsets under an additive model: p=0.0042, OR 1.23 (95% CI 1.07 to 1.41) and p=0.0065 OR 1.19 (95% CI 1.05 to 1.36), respectively. Logistic regression analysis showed an additive effect of IRF5 rs2004640, STAT4 rs7574865 and NLRP1 rs8182352 risk alleles on SSc-related FA. Conclusions Our results establish NLRP1 as a new genetic susceptibility factor for SSc-related pulmonary fibrosis and anti-topoisomerase-positive SSc phenotypes. This provides new insights into the pathogenesis of SSc, underlining the potential role of innate immunity in particular in the FA-positive SSc subphenotype, which represents a severe subset of the disease.

Variants of the MATP/SLC45A2 gene are protective for melanoma in the French population.

In this study, we investigated whether variants in three key pigmentation genes—MC1R, MATP/SLC45A2, and OCA2—were involved in melanoma predisposition. A cohort comprising 1,019 melanoma patients (MelanCohort) and 1,466 Caucasian controls without skin cancers were studied. A total of 10 polymorphisms, including five functional MC1R alleles (p.Asp84Glu, p.Arg142His, p.Arg151Cys, p.Arg160Trp, and p.Asp294His), two nonsynonymous SLC45A2 variants (p.Phe374Leu and p.Glu272Lys), and three intronic OCA2 variants previously shown to be strongly associated with eye color (rs7495174 T>C, rs4778241 G>T, and rs4778138 T>C) were genotyped. As expected, MC1R variants were closely associated with melanoma risk (P value <2.20.10−16; odds ratio [OR]=2.29 [95% confidence interval, CI=1.85–2.82 and OR=3.3 [95% CI=2.00–5.45], for the presence of one or two variants, respectively). Interestingly, the SLC45A2 variant p.Phe374Leu was significantly and strongly protective for melanoma (P‐value=2.12.10−15; OR=0.35 [95% CI=0.26–0.46] and OR=0.32 [95% CI=0.24–0.43], considering the genotypes Phe/Leu and Leu/Leu, respectively). MC1R and SLC45A2 variants had additive effects on melanoma risk, and after adjusting for pigmentation characteristics, the risk was persistent, even though both genes had a strong impact on pigmentation. Future studies may show whether genetic information could provide a useful complement to physical examination in predicting melanoma risk. Hum Mutat 0,1–7, 2008.