Integrative meta-analysis of differential gene expression in acute myeloid leukemia

Acute pancreatitis as the initial manifestation of acute myeloid leukemia with chromosome 16 rearrangements

Acute pancreatitis is an acute inflammatory process of the pancreas that is becoming an increasingly common clinical issue. The most frequent underlying etiologies include gallstones and chronic alcohol use, which account for more than two-thirds of cases. We recently experienced a rare case of acute myeloid leukemia (AML) presenting with recurrent acute pancreatitis, which we later discovered was caused by diffusely infiltrating extramedullary sarcoma in the pancreas. Comprehensive analysis of previous cases of AML presenting as acute pancreatitis suggested involvement of cytogenetic alterations in chromosome 16 in its pathogenesis. Further improvement in management of acute pancreatitis is needed, and clinicians should note that this occasionally fatal condition can be the initial and only manifestation of AML. In practice, prompt initiation of intensive chemotherapy is critical for treating such cases of AML-induced acute pancreatitis.

Trisomy 8 in acute myeloid leukemia

Introduction: Trisomy 8 is one of the most common cytogenetic alterations in acute myeloid leukemia (AML), with a frequency between 10% and 15%.Areas covered: The authors summarize the latest research regarding biological, translational and clinical aspects of trisomy 8 in AML.Expert opinion: Trisomy 8 can be found together with other karyotypes, although it also occurs as a sole aberration. The last decade's research has brought attention to molecular genetic alterations as strong contributors of leukemogenesis. AML with trisomy 8 seems to be associated with mutations in DNA methylation genes, spliceosome complex genes, and myeloid transcription factor genes, and these alterations probably have stronger implication for leukemic pathogenesis, treatment and hence prognosis, than the existence of trisomy 8 itself. Especially mutations in the RUNX1 and ASXL1 genes occur in high frequencies, and search for such mutations should be mandatory part of the diagnostic workup. AML with trisomy 8 is classified as intermediate-risk AML after recent European Leukemia Net (ELN) classification, and hence allogenic hematopoietic stem cell transplantation (Allo-HSCT) should be consider as consolidation therapy for this patient group.Trisomy 8 is frequently occurring in AML, although future molecular genetic workup should be performed, to optimize the diagnosis and treatment of these patients.

A CD33 Antigen-Targeted AAV6 Vector Expressing an Inducible Caspase-9 Suicide Gene Is Therapeutic in a Xenotransplantation Model of Acute Myeloid Leukemia

Current chemotherapeutic regimens for acute myeloid leukemia (AML) have been modestly effective in patients and are associated with poor long-term survival (<30% at 5 years). Viral vector-based suicide gene therapy is an attractive option, if these vectors can target the AML cells with high specificity and efficiency. In this study, we have developed a receptor-specific adeno-associated virus (AAV) based vector to target the CD33 antigen which is overexpressed in leukemic cells. A targeting peptide was rationally designed from the antigen-binding regions of a CD33 monoclonal antibody. This peptide was further expressed on the capsid of the AAV6 vector, since this serotype was most efficient among AAV1-rh10 vectors to infect the pro-monocytic, human myeloid leukemia cells (U937). AAV6-CD33 vectors expressing a suicide gene, the inducible caspase 9 (iCasp9), and its prodrug AP20187 significantly reduced (∼59%) the viability of U937 cells. To further test its efficacy and specificity in vivo, AAV6-CD33 vectors were administered into a xenotransplantation model of AML in zebrafish through systemic delivery. We observed a significant antileukemic effect with AAV6-CD33 vectors, with a markedly higher survival (100% for AAV6-CD33 vectors vs 15% for mock-treated) and a higher number of TUNEL positive apoptotic cells after systemic vector delivery. Taken together, our work demonstrates the efficacy and translational potential of CD33-targeted AAV6 vectors for cytotoxic gene therapy in AML.

Multi-study reanalysis of 2,213 acute myeloid leukemia patients reveals age- and sex-dependent gene expression signatures

In 2019 it is estimated that more than 21,000 new acute myeloid leukemia (AML) patients will be diagnosed in the United States, and nearly 11,000 are expected to die from the disease. AML is primarily diagnosed among the elderly (median 68 years old at diagnosis). Prognoses have significantly improved for younger patients, but as much as 70% of patients over 60 years old will die within a year of diagnosis. In this study, we conducted a reanalysis of 2,213 acute myeloid leukemia patients compared to 548 healthy individuals, using curated publicly available microarray gene expression data. We carried out an analysis of normalized batch corrected data, using a linear model that included considerations for disease, age, sex, and tissue. We identified 974 differentially expressed probe sets and 4 significant pathways associated with AML. Additionally, we identified 375 age- and 70 sex-related probe set expression signatures relevant to AML. Finally, we trained a k nearest neighbors model to classify AML and healthy subjects with 90.9% accuracy. Our findings provide a new reanalysis of public datasets, that enabled the identification of new gene sets relevant to AML that can potentially be used in future experiments and possible stratified disease diagnostics.

Not Only Mutations Matter: Molecular Picture of Acute Myeloid Leukemia Emerging from Transcriptome Studies

The last two decades of genome-scale research revealed a complex molecular picture of acute myeloid leukemia (AML). On the one hand, a number of mutations were discovered and associated with AML diagnosis and prognosis; some of them were introduced into diagnostic tests. On the other hand, transcriptome studies, which preceded AML exome and genome sequencing, remained poorly translated into clinics. Nevertheless, gene expression studies significantly contributed to the elucidation of AML pathogenesis and indicated potential therapeutic directions. The power of transcriptomic approach lies in its comprehensiveness; we can observe how genome manifests its function in a particular type of cells and follow many genes in one test. Moreover, gene expression measurement can be combined with mutation detection, as high-impact mutations are often present in transcripts. This review sums up 20 years of transcriptome research devoted to AML. Gene expression profiling (GEP) revealed signatures distinctive for selected AML subtypes and uncovered the additional within-subtype heterogeneity. The results were particularly valuable in the case of AML with normal karyotype which concerns up to 50% of AML cases. With the use of GEP, new classes of the disease were identified and prognostic predictors were proposed. A plenty of genes were detected as overexpressed in AML when compared to healthy control, including KIT, BAALC, ERG, MN1, CDX2, WT1, PRAME, and HOX genes. High expression of these genes constitutes usually an unfavorable prognostic factor. Upregulation of FLT3 and NPM1 genes, independent on their mutation status, was also reported in AML and correlated with poor outcome. However, transcriptome is not limited to the protein-coding genes; other types of RNA molecules exist in a cell and regulate genome function. It was shown that microRNA (miRNA) profiles differentiated AML groups and predicted outcome not worse than protein-coding gene profiles. For example, upregulation of miR-10a, miR-10b, and miR-196b and downregulation of miR-192 were found as typical of AML with NPM1 mutation whereas overexpression of miR-155 was associated with FLT3-internal tandem duplication (FLT3-ITD). Development of high-throughput technologies and microarray replacement by next generation sequencing (RNA-seq) enabled uncovering a real variety of leukemic cell transcriptomes, reflected by gene fusions, chimeric RNAs, alternatively spliced transcripts, miRNAs, piRNAs, long noncoding RNAs (lncRNAs), and their special type, circular RNAs. Many of them can be considered as AML biomarkers and potential therapeutic targets. The relations between particular RNA puzzles and other components of leukemic cells and their microenvironment, such as exosomes, are now under investigation. Hopefully, the results of this research will shed the light on these aspects of AML pathogenesis which are still not completely understood.

GSMA: an approach to identify robust global and test Gene Signatures using Meta-Analysis

MOTIVATION

Recent advances in biomedical research have made massive amount of transcriptomic data available in public repositories from different sources. Due to the heterogeneity present in the individual experiments, identifying reproducible biomarkers for a given disease from multiple independent studies has become a major challenge. The widely used meta-analysis approaches, such as Fisher's method, Stouffer's method, minP and maxP, have at least two major limitations: (i) they are sensitive to outliers, and (ii) they perform only one statistical test for each individual study, and hence do not fully utilize the potential sample size to gain statistical power.

RESULTS

Here, we propose a gene-level meta-analysis framework that overcomes these limitations and identifies a gene signature that is reliable and reproducible across multiple independent studies of a given disease. The approach provides a comprehensive global signature that can be used to understand the underlying biological phenomena, and a smaller test signature that can be used to classify future samples of a given disease. We demonstrate the utility of the framework by constructing disease signatures for influenza and Alzheimer's disease using nine datasets including 1108 individuals. These signatures are then validated on 12 independent datasets including 912 individuals. The results indicate that the proposed approach performs better than the majority of the existing meta-analysis approaches in terms of both sensitivity as well as specificity. The proposed signatures could be further used in diagnosis, prognosis and identification of therapeutic targets.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Age-specific biological and molecular profiling distinguishes paediatric from adult acute myeloid leukaemias

Acute myeloid leukaemia (AML) affects children and adults of all ages. AML remains one of the major causes of death in children with cancer and for children with AML relapse is the most common cause of death. Here, by modelling AML in vivo we demonstrate that AML is discriminated by the age of the cell of origin. Young cells give rise to myeloid, lymphoid or mixed phenotype acute leukaemia, whereas adult cells give rise exclusively to AML, with a shorter latency. Unlike adult, young AML cells do not remodel the bone marrow stroma. Transcriptional analysis distinguishes young AML by the upregulation of immune pathways. Analysis of human paediatric AML samples recapitulates a paediatric immune cell interaction gene signature, highlighting two genes, RGS10 and FAM26F as prognostically significant. This work advances our understanding of paediatric AML biology, and provides murine models that offer the potential for developing paediatric specific therapeutic strategies.

Acute myeloid leukaemia (AML) affects people of all ages. Here, the authors model AML in vivo and demonstrate that the age of the cell of origin impacts leukaemia development and the genetic signature where adult cells of origin give rise exclusively to AML and young cells of origin give rise to myeloid, lymphoid or mixed phenotype acute leukaemia.

A 4-gene expression prognostic signature might guide post-remission therapy in patients with intermediate-risk cytogenetic acute myeloid leukemia

In intermediate-risk cytogenetic acute myeloid leukemia (IRC-AML) patients, novel biomarkers to guide post-remission therapy are needed. We analyzed with high-density arrays 40 IRC-AML patients who received a non-allogeneic hematopoietic stem-cell transplantation-based post-remission therapy, and identified a signature that correlated with early relapse. Subsequently, we analyzed selected 187 genes in 49 additional IRC-AML patients by RT-PCR. BAALC, MN1, SPARC and HOPX overexpression correlated to refractoriness. BAALC or ALDH2 overexpression correlated to shorter overall survival (OS) (5-year OS: 33 ± 8.6% vs. 73.7 ± 10.1%, p = .006; 32 ± 9.3% vs. 66.4 ± 9.7%, p = .016), whereas GPR44 or TP53INP1 overexpression correlated to longer survival (5-year OS: 66.7 ± 10.3% vs. 35.4 ± 9.1%, p = .04; 58.3 ± 8.2% vs. 23.1 ± 11.7%, p = .029). A risk-score combining these four genes expression distinguished low-risk and high-risk patients (5-year OS: 79 ± 9% vs. 30 ± 8%, respectively; p = .001) in our cohort and in an independent set of patients from a public repository. Our 4-gene signature may add prognostic information and guide post-remission treatment in IRC-AML patients.

A four-gene LincRNA expression signature predicts risk in multiple cohorts of acute myeloid leukemia patients

Prognostic gene expression signatures have been proposed as clinical tools to clarify therapeutic options in acute myeloid leukemia (AML). However, these signatures rely on measuring large numbers of genes and often perform poorly when applied to independent cohorts or those with older patients. Long intergenic non-coding RNAs (lincRNAs) are emerging as important regulators of cell identity and oncogenesis, but knowledge of their utility as prognostic markers in AML is limited. Here we analyze transcriptomic data from multiple cohorts of clinically annotated AML patients and report that (i) microarrays designed for coding gene expression can be repurposed to yield robust lincRNA expression data, (ii) some lincRNA genes are located in close proximity to hematopoietic coding genes and show strong expression correlations in AML, (iii) lincRNA gene expression patterns distinguish cytogenetic and molecular subtypes of AML, (iv) lincRNA signatures composed of three or four genes are independent predictors of clinical outcome and further dichotomize survival in European Leukemia Net (ELN) risk groups and (v) an analytical tool based on logistic regression analysis of quantitative PCR measurement of four lincRNA genes (LINC4) can be used to determine risk in AML.

Large-scale gene network analysis reveals the significance of extracellular matrix pathway and homeobox genes in acute myeloid leukemia: an introduction to the Pigengene package and its applications

BACKGROUND

The distinct types of hematological malignancies have different biological mechanisms and prognoses. For instance, myelodysplastic syndrome (MDS) is generally indolent and low risk; however, it may transform into acute myeloid leukemia (AML), which is much more aggressive.

METHODS

We develop a novel network analysis approach that uses expression of eigengenes to delineate the biological differences between these two diseases.

RESULTS

We find that specific genes in the extracellular matrix pathway are underexpressed in AML. We validate this finding in three ways: (a) We train our model on a microarray dataset of 364 cases and test it on an RNA Seq dataset of 74 cases. Our model showed 95% sensitivity and 86% specificity in the training dataset and showed 98% sensitivity and 91% specificity in the test dataset. This confirms that the identified biological signatures are independent from the expression profiling technology and independent from the training dataset. (b) Immunocytochemistry confirms that MMP9, an exemplar protein in the extracellular matrix, is underexpressed in AML. (c) MMP9 is hypermethylated in the majority of AML cases (n=194, Welch's t-test p-value <10^-138), which complies with its low expression in AML. Our novel network analysis approach is generalizable and useful in studying other complex diseases (e.g., breast cancer prognosis). We implement our methodology in the Pigengene software package, which is publicly available through Bioconductor.

CONCLUSIONS

Eigengenes define informative biological signatures that are robust with respect to expression profiling technology. These signatures provide valuable information about the underlying biology of diseases, and they are useful in predicting diagnosis and prognosis.

Independent validation test of the vote-counting strategy used to rank biomarkers from published studies

AIM

Vote counting is frequently used in meta-analyses to rank biomarker candidates, but to our knowledge, there have been no independent assessments of its validity. Here, we used predictions from a recent meta-analysis to determine how well number of supporting studies, combined sample size and mean fold change performed as vote-counting strategy criteria.

MATERIALS & METHODS

Fifty miRNAs previously ranked for their ability to distinguish lung cancer tissue from normal were assayed by RT-qPCR using 45 paired tumor-normal samples.

RESULTS

Number of supporting studies predicted biomarker performance (p = 0.0006; r = 0.44), but sample size and fold change did not (p > 0.2).

CONCLUSION

Despite limitations, counting the number supporting studies appears to be an effective criterion for ranking biomarkers. Predictions based on sample size and fold change provided little added value. External validation studies should be conducted to establish the performance characteristics of strategies used to rank biomarkers.

Meta-analysis of gene expression signatures reveals hidden links among diverse biological processes in Arabidopsis

The model plant Arabidopsis has been well-studied using high-throughput genomics technologies, which usually generate lists of differentially expressed genes under various conditions. Our group recently collected 1065 gene lists from 397 gene expression studies as a knowledgebase for pathway analysis. Here we systematically analyzed these gene lists by computing overlaps in all-vs.-all comparisons. We identified 16,261 statistically significant overlaps, represented by an undirected network in which nodes correspond to gene lists and edges indicate significant overlaps. The network highlights the correlation across the gene expression signatures of the diverse biological processes. We also partitioned the main network into 20 sub-networks, representing groups of highly similar expression signatures. These are common sets of genes that were co-regulated under different treatments or conditions and are often related to specific biological themes. Overall, our result suggests that diverse gene expression signatures are highly interconnected in a modular fashion.

Identification of the RNA recognition element of the RBPMS family of RNA-binding proteins and their transcriptome-wide mRNA targets

Recent studies implicated the RNA-binding protein with multiple splicing (RBPMS) family of proteins in oocyte, retinal ganglion cell, heart, and gastrointestinal smooth muscle development. These RNA-binding proteins contain a single RNA recognition motif (RRM), and their targets and molecular function have not yet been identified. We defined transcriptome-wide RNA targets using photoactivatable-ribonucleoside-enhanced crosslinking and immunoprecipitation (PAR-CLIP) in HEK293 cells, revealing exonic mature and intronic pre-mRNA binding sites, in agreement with the nuclear and cytoplasmic localization of the proteins. Computational and biochemical approaches defined the RNA recognition element (RRE) as a tandem CAC trinucleotide motif separated by a variable spacer region. Similar to other mRNA-binding proteins, RBPMS family of proteins relocalized to cytoplasmic stress granules under oxidative stress conditions suggestive of a support function for mRNA localization in large and/or multinucleated cells where it is preferentially expressed.

Constitutive IRF8 expression inhibits AML by activation of repressed immune response signaling

Myeloid differentiation is blocked in acute myeloid leukemia (AML), but the molecular mechanisms are not well characterized. Meningioma 1 (MN1) is overexpressed in AML patients and confers resistance to all-trans retinoic acid-induced differentiation. To understand the role of MN1 as a transcriptional regulator in myeloid differentiation, we fused transcriptional activation (VP16) or repression (M33) domains with MN1 and characterized these cells in vivo. Transcriptional activation of MN1 target genes induced myeloproliferative disease with long latency and differentiation potential to mature neutrophils. A large proportion of differentially expressed genes between leukemic MN1 and differentiation-permissive MN1VP16 cells belonged to the immune response pathway like interferon-response factor (Irf) 8 and Ccl9. As MN1 is a cofactor of MEIS1 and retinoic acid receptor alpha (RARA), we compared chromatin occupancy between these genes. Immune response genes that were upregulated in MN1VP16 cells were co-targeted by MN1 and MEIS1, but not RARA, suggesting that myeloid differentiation is blocked through transcriptional repression of shared target genes of MN1 and MEIS1. Constitutive expression of Irf8 or its target gene Ccl9 identified these genes as potent inhibitors of murine and human leukemias in vivo. Our data show that MN1 prevents activation of the immune response pathway, and suggest restoration of IRF8 signaling as therapeutic target in AML.

Expression profiling of leukemia patients: key lessons and future directions

Gene expression profiling (GEP) is a well-established indispensable tool used to study hematologic malignancies, including leukemias. Here, we summarize the insights into the molecular basis of leukemias obtained by means of GEP, focusing especially on acute myeloid leukemia (AML), one of the first diseases to be extensively studied by GEP. Profiling mRNA and microRNA expression are discussed in view of their applicability to class prediction, class discovery, and comparison, as well as outcome prediction, and special attention is paid to the recent advances in our understanding of the role of alternative RNA splicing in AML. In addition to microarray-based GEP approaches, over the last few years RNA sequencing based on next-generation sequencing technology is gaining wider recognition as an advanced tool for transcriptome profiling. Therefore, the advantages of RNA sequencing-based GEP and its current and potential implications in AML are discussed. Finally, we also highlight recent efforts to integrate already available and newly acquired omics data sets so that a more precise understanding of AML biology and clinical behavior can be achieved, which ultimately will contribute to further refine leukemia management.

Novel t(7;10)(p22;p24) along with NPM1 mutation in patient with relapsed acute myeloid leukemia

Chromosomal abnormalities/genetic mutations associated with hematological malignancies alter the structure and function of genes controlling cell proliferation and differentiation through multiple and complex pathways, resulting different clinical outcomes. This is a case study of a lady presented with acute myeloid leukemia (AML M1) at our center who relapsed 10 years after the induction therapy. Cytogenetic and molecular analyses were performed in this case at the time of relapse to find out the chromosomal abnormalities and genetic abnormalities like FMS-like tyrosine kinase (FLT3) and nucleophosmin (NPM1) mutation. The cytogenetic analysis of bone marrow established a novel translocation t(7;10) (p22;q24) in 100% of the cells analyzed. Phytohaemagglutinin (PHA)-stimulated blood culture also revealed the same abnormality. Apart from this, the molecular analysis showed NPM1 exon 12 (hot-spot) mutation in this patient. This was the first report of novel chromosomal translocation in this subset of AML in which a new translocation along with NPM1 mutation was discussed.

Analyzing illumina gene expression microarray data from different tissues: methodological aspects of data analysis in the metaxpress consortium

Microarray profiling of gene expression is widely applied in molecular biology and functional genomics. Experimental and technical variations make meta-analysis of different studies challenging. In a total of 3358 samples, all from German population-based cohorts, we investigated the effect of data preprocessing and the variability due to sample processing in whole blood cell and blood monocyte gene expression data, measured on the Illumina HumanHT-12 v3 BeadChip array.

Gene expression signal intensities were similar after applying the log₂ or the variance-stabilizing transformation. In all cohorts, the first principal component (PC) explained more than 95% of the total variation. Technical factors substantially influenced signal intensity values, especially the Illumina chip assignment (33–48% of the variance), the RNA amplification batch (12–24%), the RNA isolation batch (16%), and the sample storage time, in particular the time between blood donation and RNA isolation for the whole blood cell samples (2–3%), and the time between RNA isolation and amplification for the monocyte samples (2%). White blood cell composition parameters were the strongest biological factors influencing the expression signal intensities in the whole blood cell samples (3%), followed by sex (1–2%) in both sample types. Known single nucleotide polymorphisms (SNPs) were located in 38% of the analyzed probe sequences and 4% of them included common SNPs (minor allele frequency >5%). Out of the tested SNPs, 1.4% significantly modified the probe-specific expression signals (Bonferroni corrected p-value<0.05), but in almost half of these events the signal intensities were even increased despite the occurrence of the mismatch. Thus, the vast majority of SNPs within probes had no significant effect on hybridization efficiency.

In summary, adjustment for a few selected technical factors greatly improved reliability of gene expression analyses. Such adjustments are particularly required for meta-analyses.

Using bioinformatic approaches to identify pathways targeted by human leukemogens

We have applied bioinformatic approaches to identify pathways common to chemical leukemogens and to determine whether leukemogens could be distinguished from non-leukemogenic carcinogens. From all known and probable carcinogens classified by IARC and NTP, we identified 35 carcinogens that were associated with leukemia risk in human studies and 16 non-leukemogenic carcinogens. Using data on gene/protein targets available in the Comparative Toxicogenomics Database (CTD) for 29 of the leukemogens and 11 of the non-leukemogenic carcinogens, we analyzed for enrichment of all 250 human biochemical pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The top pathways targeted by the leukemogens included metabolism of xenobiotics by cytochrome P450, glutathione metabolism, neurotrophin signaling pathway, apoptosis, MAPK signaling, Toll-like receptor signaling and various cancer pathways. The 29 leukemogens formed 18 distinct clusters comprising 1 to 3 chemicals that did not correlate with known mechanism of action or with structural similarity as determined by 2D Tanimoto coefficients in the PubChem database. Unsupervised clustering and one-class support vector machines, based on the pathway data, were unable to distinguish the 29 leukemogens from 11 non-leukemogenic known and probable IARC carcinogens. However, using two-class random forests to estimate leukemogen and non-leukemogen patterns, we estimated a 76% chance of distinguishing a random leukemogen/non-leukemogen pair from each other.

FLT3-ITD-associated gene-expression signatures in NPM1-mutated cytogenetically normal acute myeloid leukemia

Concomitance of the FLT3-ITD mutation is associated with poor prognosis in NPM1-mutated cytogenetically normal acute myeloid leukemia (CN-AML) patients, and precise studies on its role in leukemogenesis are needed; these may be elucidated at the molecular level by gene express profiling. In the present study, we built a gene-expression-based classifier using prediction analysis of microarray to characterize the FLT3-ITD signature in NPM1-mutated CN-AML patients, which comprised 10 annotated genes, and demonstrated an overall accuracy of 83.8 % in cross-validation. To characterize the signature in another way, differential expression was revealed for 34 genes by class comparison, and the up-regulation of LAPTM4B and MIR155HG was validated by quantitative RT-PCR in our small cohort of NPM1-mutated CN-AML samples, which appeared to be associated with this specific subtype. The 10-gene classifier and differentially expressed genes identified in this study indicate a potential utility for risk-assessed treatment stratification, and suggest new therapeutic targets for these high-risk AML patients.

Exact statistical tests for the intersection of independent lists of genes

Public data repositories have enabled researchers to compare results across multiple genomic studies in order to replicate findings. A common approach is to first rank genes according to an hypothesis of interest within each study. Then, lists of the top-ranked genes within each study are compared across studies. Genes recaptured as highly ranked (usually above some threshold) in multiple studies are considered to be significant. However, this comparison strategy often remains informal, in that Type I error and false discovery rate are usually uncontrolled. In this paper, we formalize an inferential strategy for this kind of list-intersection discovery test. We show how to compute a p-value associated with a `recaptured' set of genes, using a closed-form Poisson approximation to the distribution of the size of the recaptured set. The distribution of the test statistic depends on the rank threshold and the number of studies within which a gene must be recaptured. We use a Poisson approximation to investigate operating characteristics of the test. We give practical guidance on how to design a bioinformatic list-intersection study with prespecified control of Type I error (at the set level) and false discovery rate (at the gene level). We show how choice of test parameters will affect the expected proportion of significant genes identified. We present a strategy for identifying optimal choice of parameters, depending on the particular alternative hypothesis which might hold. We illustrate our methods using prostate cancer gene-expression datasets from the curated Oncomine database.

Robust rank aggregation for gene list integration and meta-analysis

MOTIVATION

The continued progress in developing technological platforms, availability of many published experimental datasets, as well as different statistical methods to analyze those data have allowed approaching the same research question using various methods simultaneously. To get the best out of all these alternatives, we need to integrate their results in an unbiased manner. Prioritized gene lists are a common result presentation method in genomic data analysis applications. Thus, the rank aggregation methods can become a useful and general solution for the integration task.

RESULTS

Standard rank aggregation methods are often ill-suited for biological settings where the gene lists are inherently noisy. As a remedy, we propose a novel robust rank aggregation (RRA) method. Our method detects genes that are ranked consistently better than expected under null hypothesis of uncorrelated inputs and assigns a significance score for each gene. The underlying probabilistic model makes the algorithm parameter free and robust to outliers, noise and errors. Significance scores also provide a rigorous way to keep only the statistically relevant genes in the final list. These properties make our approach robust and compelling for many settings.

AVAILABILITY

All the methods are implemented as a GNU R package RobustRankAggreg, freely available at the Comprehensive R Archive Network http://cran.r-project.org/.

Acute myeloid leukemia with the t(8;21) translocation: clinical consequences and biological implications

The t(8;21) abnormality occurs in a minority of acute myeloid leukemia (AML) patients. The translocation results in an in-frame fusion of two genes, resulting in a fusion protein of one N-terminal domain from the AML1 gene and four C-terminal domains from the ETO gene. This protein has multiple effects on the regulation of the proliferation, the differentiation, and the viability of leukemic cells. The translocation can be detected as the only genetic abnormality or as part of more complex abnormalities. If t(8;21) is detected in a patient with bone marrow pathology, the diagnosis AML can be made based on this abnormality alone. t(8;21) is usually associated with a good prognosis. Whether the detection of the fusion gene can be used for evaluation of minimal residual disease and risk of leukemia relapse remains to be clarified. To conclude, detection of t(8;21) is essential for optimal handling of these patients as it has both diagnostic, prognostic, and therapeutic implications.

Identification of predictive markers of cytarabine response in AML by integrative analysis of gene-expression profiles with multiple phenotypes

AIM

To identify gene-expression signatures predicting cytarabine response by an integrative analysis of multiple clinical and pharmacological end points in acute myeloid leukemia (AML) patients.

MATERIALS & METHODS

We performed an integrated analysis to associate the gene expression of diagnostic bone marrow blasts from acute myeloid leukemia (AML) patients treated in the discovery set (AML97; n = 42) and in the independent validation set (AML02; n = 46) with multiple clinical and pharmacological end points. Based on prior biological knowledge, we defined a gene to show a therapeutically beneficial (detrimental) pattern of association of its expression positively (negatively) correlated with favorable phenotypes such as intracellular cytarabine 5´-triphosphate levels, morphological response and event-free survival, and negatively (positively) correlated with unfavorable end points such as post-cytarabine DNA synthesis levels, minimal residual disease and cytarabine LC(50).

RESULTS

We identified 240 probe sets predicting a therapeutically beneficial pattern and 97 predicting detrimental pattern (p ≤ 0.005) in the discovery set. Of these, 60 were confirmed in the independent validation set. The validated probe sets correspond to genes involved in PIK3/PTEN/AKT/mTOR signaling, G-protein-coupled receptor signaling and leukemogenesis. This suggests that targeting these pathways as potential pharmacogenomic and therapeutic candidates could be useful for improving treatment outcomes in AML.

CONCLUSION

This study illustrates the power of integrated data analysis of genomic data as well as multiple clinical and pharmacologic end points in the identification of genes and pathways of biological relevance.

Ontology-based meta-analysis of global collections of high-throughput public data

Background

The investigation of the interconnections between the molecular and genetic events that govern biological systems is essential if we are to understand the development of disease and design effective novel treatments. Microarray and next-generation sequencing technologies have the potential to provide this information. However, taking full advantage of these approaches requires that biological connections be made across large quantities of highly heterogeneous genomic datasets. Leveraging the increasingly huge quantities of genomic data in the public domain is fast becoming one of the key challenges in the research community today.

Methodology/Results

We have developed a novel data mining framework that enables researchers to use this growing collection of public high-throughput data to investigate any set of genes or proteins. The connectivity between molecular states across thousands of heterogeneous datasets from microarrays and other genomic platforms is determined through a combination of rank-based enrichment statistics, meta-analyses, and biomedical ontologies. We address data quality concerns through dataset replication and meta-analysis and ensure that the majority of the findings are derived using multiple lines of evidence. As an example of our strategy and the utility of this framework, we apply our data mining approach to explore the biology of brown fat within the context of the thousands of publicly available gene expression datasets.

Conclusions

Our work presents a practical strategy for organizing, mining, and correlating global collections of large-scale genomic data to explore normal and disease biology. Using a hypothesis-free approach, we demonstrate how a data-driven analysis across very large collections of genomic data can reveal novel discoveries and evidence to support existing hypothesis.