1
|
Novianti PW, Jong VL, Roes KCB, Eijkemans MJC. Meta-analysis approach as a gene selection method in class prediction: does it improve model performance? A case study in acute myeloid leukemia. BMC Bioinformatics 2017; 18:210. [PMID: 28399794 PMCID: PMC5387259 DOI: 10.1186/s12859-017-1619-7] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.6] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/05/2016] [Accepted: 03/30/2017] [Indexed: 12/22/2022] Open
Abstract
BACKGROUND Aggregating gene expression data across experiments via meta-analysis is expected to increase the precision of the effect estimates and to increase the statistical power to detect a certain fold change. This study evaluates the potential benefit of using a meta-analysis approach as a gene selection method prior to predictive modeling in gene expression data. RESULTS Six raw datasets from different gene expression experiments in acute myeloid leukemia (AML) and 11 different classification methods were used to build classification models to classify samples as either AML or healthy control. First, the classification models were trained on gene expression data from single experiments using conventional supervised variable selection and externally validated with the other five gene expression datasets (referred to as the individual-classification approach). Next, gene selection was performed through meta-analysis on four datasets, and predictive models were trained with the selected genes on the fifth dataset and validated on the sixth dataset. For some datasets, gene selection through meta-analysis helped classification models to achieve higher performance as compared to predictive modeling based on a single dataset; but for others, there was no major improvement. Synthetic datasets were generated from nine simulation scenarios. The effect of sample size, fold change and pairwise correlation between differentially expressed (DE) genes on the difference between MA- and individual-classification model was evaluated. The fold change and pairwise correlation significantly contributed to the difference in performance between the two methods. The gene selection via meta-analysis approach was more effective when it was conducted using a set of data with low fold change and high pairwise correlation on the DE genes. CONCLUSION Gene selection through meta-analysis on previously published studies potentially improves the performance of a predictive model on a given gene expression data.
Collapse
Affiliation(s)
- Putri W. Novianti
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, 3508 GA Utrecht, The Netherlands
- Department of Epidemiology and Biostatistics, VU University medical center, Amsterdam, The Netherlands
- Department of Pathology, VU University medical center, Amsterdam, The Netherlands
| | - Victor L. Jong
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, 3508 GA Utrecht, The Netherlands
- Viroscience Laboratory, Erasmus Medical Center Rotterdam, 3015 CE Rotterdam, The Netherlands
| | - Kit C. B. Roes
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, 3508 GA Utrecht, The Netherlands
| | - Marinus J. C. Eijkemans
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, 3508 GA Utrecht, The Netherlands
| |
Collapse
|
2
|
Martin-Subero M, Diez-Quevedo C. Mental disorders in HIV/HCV coinfected patients under antiviral treatment for hepatitis C. Psychiatry Res 2016; 246:173-181. [PMID: 27718466 DOI: 10.1016/j.psychres.2016.09.041] [Citation(s) in RCA: 7] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 01/15/2016] [Revised: 07/27/2016] [Accepted: 09/24/2016] [Indexed: 12/14/2022]
Abstract
This paper aims to review the epidemiology and management of mental disorders in human immunodeficiency virus (HIV)/hepatitis C virus (HCV) coinfected patients, the need for antiviral therapy in this specific population, and current treatment strategies for HIV/HCV patients with psychiatric and/or substance use disorders. This is a narrative review. Data was sourced from electronic databases and was not limited by language or date of publication. HIV infection has become a survivable chronic illness. Prevalence of HCV infection among HIV-infected patients is high ranging from 50% to 90%. Patients with psychiatric diseases have also an increased risk for HIV/HCV coinfection. The most effective strategy to decrease HCV-related morbidity and mortality in coinfection is to achieve viral eradication. Although psychiatric symptoms often appear during antiviral treatment and may be associated with the use of interferon-alpha (IFN-α), recent evidence suggests that many patients with comorbid mental and substance use disorders can be treated safely. Recent data indicate that IFNα-induced psychiatric side effects have a similar prevalence in HIV/HCV coinfected patients than in monoinfected patients and they can be managed and even prevented successfully with psychopharmacological strategies in the frame of a multidisciplinary team. New antivirals offer INF-free therapies for this specific population.
Collapse
Affiliation(s)
- Marta Martin-Subero
- FIDMAG Research Foundation, Barcelona, Spain; Department of Psychiatry and Forensic Medicine, Universitat Autònoma de Barcelona, Badalona, Spain.
| | - Crisanto Diez-Quevedo
- Department of Psychiatry and Forensic Medicine, Universitat Autònoma de Barcelona, Badalona, Spain; Department of Psychiatry, Hospital Universitari Germans Trias i Pujol, Badalona, Spain
| |
Collapse
|
3
|
Katsounas A, Wilting KR, Lempicki RA, Schlaak JF, Gerken G. Microarrays-Enabled Hypothesis Generation: The Suspect Role of FNBP-1 in Neuropsychiatric Pathogenesis Associated with HIV and/or HCV Infection. JOURNAL OF AIDS & CLINICAL RESEARCH 2016; 7:641. [PMID: 28255515 PMCID: PMC5330367 DOI: 10.4172/2155-6113.1000641] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Indexed: 11/12/2022]
Abstract
OBJECTIVE The spectrum of neuropsychiatric illness (NI) associated with the Human Immunodeficiency Virus (HIV) and/or the Hepatitis C Virus (HCV) is far reaching and significantly impacts the clinical presentation and outcome of infected persons; however, the etiological and pathophysiological background remains partially understood. The present work was aimed to investigate the potential significance of formin binding protein 1 (FNBP-1)-dependent pathways in NI-pathogenesis by elaborating on previous microarray-based research in HIV and/or HCV-infected patients receiving interferon-α (IFN-α) immunotherapy via a rigorous data mining procedure. METHODS Using microarray data of peripheral whole blood (PB) samples obtained from HCV mono-infected persons (n=25, Affymetrix® HG-U133A_2) 12 h before and after the 1st dose of pegylated IFN-α (PegIFN-α), we re-applied the same analytical algorithm that we had developed and published in an earlier study with HIV/HCV co-infected subjects (N=28, Affymetrix® HG-U133A), in order to evaluate reproducibility of potential NI-related molecular findings in an independent cohort. RESULTS Among 28 gene expression profiles (HIV/HCV: N=9 vs. HCV: N=19) selected by applying different thresholds (a Mean Fold Difference value (MFD) in gene expression of ≥ 0.38 (log2) and/or P value from <0.05 to ≤ 0.1) FNBP-1 was identified as the only overlapping marker, which also exhibited a consistent upregulation in association with the development of NI in both cohorts. Previous functional annotation analysis had classified FNBP-1 as molecule with significant enrichment in various brain tissues (P<0.01). CONCLUSION Our current findings are strongly arguing for intensifying research into the FNBP-1-related mechanisms that may be conferring risk for or resistance to HIV- and/or HCV-related NI.
Collapse
Affiliation(s)
- A Katsounas
- Department of Gastroenterology and Hepatology, University Hospital Essen, Hufelandstrasse 55, 45147 Essen, Germany
- Laboratory of Immunopathogenesis and Bioinformatics, Leidos Biomedical Research, Inc., National Cancer Institute at Frederick, Frederick, MD 21702, USA
| | - KR Wilting
- Department for Medical Microbiology and Infection Prevention, University Medical Center Groningen, Hanzeplein 1 (9713 GZ) Groningen, the Netherlands
| | - RA Lempicki
- Laboratory of Immunopathogenesis and Bioinformatics, Leidos Biomedical Research, Inc., National Cancer Institute at Frederick, Frederick, MD 21702, USA
| | - JF Schlaak
- Evangelisches Klinikum Niederrhein gGmbH, Duisburg, Germany
| | - G Gerken
- Department of Gastroenterology and Hepatology, University Hospital Essen, Hufelandstrasse 55, 45147 Essen, Germany
| |
Collapse
|
4
|
Differences in gene expression and alterations in cell cycle of acute myeloid leukemia cell lines after treatment with JAK inhibitors. Eur J Pharmacol 2015; 765:188-97. [DOI: 10.1016/j.ejphar.2015.08.037] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/19/2015] [Revised: 08/18/2015] [Accepted: 08/19/2015] [Indexed: 12/15/2022]
|
5
|
Factors affecting the accuracy of a class prediction model in gene expression data. BMC Bioinformatics 2015; 16:199. [PMID: 26093633 PMCID: PMC4475623 DOI: 10.1186/s12859-015-0610-4] [Citation(s) in RCA: 13] [Impact Index Per Article: 1.4] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/08/2015] [Accepted: 04/30/2015] [Indexed: 01/12/2023] Open
Abstract
BACKGROUND Class prediction models have been shown to have varying performances in clinical gene expression datasets. Previous evaluation studies, mostly done in the field of cancer, showed that the accuracy of class prediction models differs from dataset to dataset and depends on the type of classification function. While a substantial amount of information is known about the characteristics of classification functions, little has been done to determine which characteristics of gene expression data have impact on the performance of a classifier. This study aims to empirically identify data characteristics that affect the predictive accuracy of classification models, outside of the field of cancer. RESULTS Datasets from twenty five studies meeting predefined inclusion and exclusion criteria were downloaded. Nine classification functions were chosen, falling within the categories: discriminant analyses or Bayes classifiers, tree based, regularization and shrinkage and nearest neighbors methods. Consequently, nine class prediction models were built for each dataset using the same procedure and their performances were evaluated by calculating their accuracies. The characteristics of each experiment were recorded, (i.e., observed disease, medical question, tissue/cell types and sample size) together with characteristics of the gene expression data, namely the number of differentially expressed genes, the fold changes and the within-class correlations. Their effects on the accuracy of a class prediction model were statistically assessed by random effects logistic regression. The number of differentially expressed genes and the average fold change had significant impact on the accuracy of a classification model and gave individual explained-variation in prediction accuracy of up to 72% and 57%, respectively. Multivariable random effects logistic regression with forward selection yielded the two aforementioned study factors and the within class correlation as factors affecting the accuracy of classification functions, explaining 91.5% of the between study variation. CONCLUSIONS We evaluated study- and data-related factors that might explain the varying performances of classification functions in non-cancerous datasets. Our results showed that the number of differentially expressed genes, the fold change, and the correlation in gene expression data significantly affect the accuracy of class prediction models.
Collapse
|
6
|
Katsounas A, Rasimas JJ, Schlaak JF, Lempicki RA, Rosenstein DL, Kottilil S. Interferon stimulated exonuclease gene 20 kDa links psychiatric events to distinct hepatitis C virus responses in human immunodeficiency virus positive patients. J Med Virol 2014; 86:1323-31. [PMID: 24782267 PMCID: PMC4114765 DOI: 10.1002/jmv.23956] [Citation(s) in RCA: 4] [Impact Index Per Article: 0.4] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Accepted: 03/18/2014] [Indexed: 01/02/2023]
Abstract
Hepatitis C Virus (HCV) infection occurs frequently in patients with preexisting mental illness. Treatment for chronic hepatitis C using interferon formulations often increases risk for neuro-psychiatric symptoms. Pegylated-Interferon-α (PegIFN-α) remains crucial for attaining sustained virologic response (SVR); however, PegIFN-α based treatment is associated with psychiatric adverse effects, which require dose reduction and/or interruption. This study's main objective was to identify genes induced by PegIFN-α and expressed in the central nervous system and immune system, which could mediate the development of psychiatric toxicity in association with antiviral outcome. Using peripheral blood mononuclear cells from Human Immunodeficiency Virus (HIV)/HCV co-infected donors (N = 28), DNA microarray analysis was performed and 21 differentially regulated genes were identified in patients with psychiatric toxicity versus those without. Using these 21 expression profiles a two-way-ANOVA was performed to select genes based on antiviral outcome and occurrence of neuro-psychiatric adverse events. Microarray analysis demonstrated that Interferon-stimulated-exonuclease-gene 20 kDa (ISG20) and Interferon-alpha-inducible-protein 27 (IFI27) were the most regulated genes (P < 0.05) between three groups that were built by combining antiviral outcome and neuro-psychiatric toxicity. Validation by bDNA assay confirmed that ISG20 expression levels were significantly associated with these outcomes (P < 0.035). Baseline levels and induction of ISG20 correlated independently with no occurrence of psychiatric adverse events and non-response to therapy (P < 0.001). Among the 21 genes that were associated with psychiatric adverse events and 20 Interferon-inducible genes (IFIGs) used as controls, only ISG20 expression was able to link PegIFN-α related neuro-psychiatric toxicity to distinct HCV-responses in patients co-infected with HIV and HCV in vivo.
Collapse
Affiliation(s)
- Antonios Katsounas
- Department of Gastroenterology and Hepatology, University Hospital Essen, Hufelandstrasse 55, 45147 Essen, Germany
| | - Joseph J. Rasimas
- Experimental Therapeutics and Pathophysiology Branch, National Institute of Mental Health, National Institutes of Health, Bethesda, MD 20892, USA
| | - Joerg F. Schlaak
- Department of Gastroenterology and Hepatology, University Hospital Essen, Hufelandstrasse 55, 45147 Essen, Germany
| | - Richard A. Lempicki
- Laboratory of Immunopathogenesis and Bioinformatics, SAIC-Frederick, Inc, National Cancer Institute at Frederick, Frederick, MD 21702, USA
| | - Donald L. Rosenstein
- Department of Psychiatry, University of North Carolina at Chapel Hill, NC 27599-7305, USA
| | - Shyam Kottilil
- Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, USA
| |
Collapse
|
7
|
Novianti PW, Roes KCB, Eijkemans MJC. Evaluation of gene expression classification studies: factors associated with classification performance. PLoS One 2014; 9:e96063. [PMID: 24770439 PMCID: PMC4000205 DOI: 10.1371/journal.pone.0096063] [Citation(s) in RCA: 10] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/01/2013] [Accepted: 04/03/2014] [Indexed: 12/22/2022] Open
Abstract
Classification methods used in microarray studies for gene expression are diverse in the way they deal with the underlying complexity of the data, as well as in the technique used to build the classification model. The MAQC II study on cancer classification problems has found that performance was affected by factors such as the classification algorithm, cross validation method, number of genes, and gene selection method. In this paper, we study the hypothesis that the disease under study significantly determines which method is optimal, and that additionally sample size, class imbalance, type of medical question (diagnostic, prognostic or treatment response), and microarray platform are potentially influential. A systematic literature review was used to extract the information from 48 published articles on non-cancer microarray classification studies. The impact of the various factors on the reported classification accuracy was analyzed through random-intercept logistic regression. The type of medical question and method of cross validation dominated the explained variation in accuracy among studies, followed by disease category and microarray platform. In total, 42% of the between study variation was explained by all the study specific and problem specific factors that we studied together.
Collapse
Affiliation(s)
- Putri W Novianti
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
| | - Kit C B Roes
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
| | - Marinus J C Eijkemans
- Biostatistics & Research Support, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands
| |
Collapse
|
8
|
Gene Expression Profiles Predict Emergence of Psychiatric Adverse Events in HIV/HCV-Coinfected Patients on Interferon-Based HCV Therapy. J Acquir Immune Defic Syndr 2012. [DOI: 10.1097/qai.0b013e318266ed8b] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/27/2022]
|