Jove
Visualize
Contact Us

Related Concept Videos

DNA Microarrays02:34

DNA Microarrays

17.6K
Microarrays are high-throughput and relatively inexpensive assays that can be automated to analyze large quantities of data at a time. They are used in genome-wide studies to compare gene or protein expression under two varied conditions, such as healthy and diseased states. Microarrays consist of glass or silica slides on which probe molecules are covalently attached through surface functionalization. Most commonly, the slides are prepared through the chemisorption of silanes to silica...
17.6K
Genome-wide Association Studies-GWAS01:11

Genome-wide Association Studies-GWAS

13.7K
Genome-wide association studies or GWAS are used to identify whether common SNPs are associated with certain diseases. Suppose specific SNPs are more frequently observed in individuals with a particular disease than those without the disease. In that case, those SNPs are said to be associated with the disease. Chi-square analysis is performed to check the probability of the allele likely to be associated with the disease.
GWAS does not require the identification of the target gene involved in...
13.7K

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

An optimized multi-task contrastive learning framework for HIFU lesion detection and segmentation.

Scientific reports·2025
Same author

Harnessing Deep Learning for Omics in an Era of COVID-19.

Omics : a journal of integrative biology·2023
Same author

A novel approach for the analysis of time-course gene expression data based on computing with words.

Journal of biomedical informatics·2021
Same author

Feature selection using genetic algorithm for breast cancer diagnosis: experiment on three different datasets.

Iranian journal of basic medical sciences·2016
Same author

Control of low-density lipoprotein concentration in the arterial wall by proportional drug-encapsulated nanoparticles.

IEEE transactions on nanobioscience·2012
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 28, 2025

Large-Scale Multi-Omics Genome-Wide Association Studies Mo-GWAS: Guidelines for Sample Preparation and Normalization
08:27

Large-Scale Multi-Omics Genome-Wide Association Studies Mo-GWAS: Guidelines for Sample Preparation and Normalization

Published on: July 27, 2021

3.7K

MS-ACGAN: A modified auxiliary classifier generative adversarial network for schizophrenia's samples augmentation

Bahareh Jahanyar1, Hamid Tabatabaee1, Alireza Rowhanimanesh2

  • 1Department of Computer Engineering, Mashhad Branch, Islamic Azad University, Mashhad, Iran.

Computers in Biology and Medicine
|June 1, 2023
PubMed
Summary

Researchers developed a new deep learning model to create synthetic gene expression data for schizophrenia research. This approach helps overcome the problem of having too few patient samples for accurate analysis. By using a specialized generative network, the team produced reliable artificial data that closely matches real-world biological patterns.

Keywords:
Adversarial generative networkConfidence intervalsData augmentationDeep learningMachine learningMicroarray gene expressionOmics dataDeep LearningGenerative Adversarial NetworksSchizophrenia GenomicsData Augmentation

Frequently Asked Questions

More Related Videos

Microarray Analysis for Saccharomyces cerevisiae
13:17

Microarray Analysis for Saccharomyces cerevisiae

Published on: April 7, 2011

13.7K
DNA Microarrays: Sample Quality Control, Array Hybridization and Scanning
09:27

DNA Microarrays: Sample Quality Control, Array Hybridization and Scanning

Published on: March 15, 2011

38.1K

Related Experiment Videos

Last Updated: Jul 28, 2025

Large-Scale Multi-Omics Genome-Wide Association Studies Mo-GWAS: Guidelines for Sample Preparation and Normalization
08:27

Large-Scale Multi-Omics Genome-Wide Association Studies Mo-GWAS: Guidelines for Sample Preparation and Normalization

Published on: July 27, 2021

3.7K
Microarray Analysis for Saccharomyces cerevisiae
13:17

Microarray Analysis for Saccharomyces cerevisiae

Published on: April 7, 2011

13.7K
DNA Microarrays: Sample Quality Control, Array Hybridization and Scanning
09:27

DNA Microarrays: Sample Quality Control, Array Hybridization and Scanning

Published on: March 15, 2011

38.1K

Area of Science:

  • Computational biology and MS-ACGAN applications in bioinformatics
  • Genomics and transcriptomics research within precision medicine

Background:

Limited access to high-quality clinical biospecimens remains a significant hurdle for advancing psychiatric research. Prior research has shown that transcriptomic datasets often suffer from small sample sizes, which restricts the performance of predictive algorithms. This gap motivated the development of synthetic data generation techniques to supplement existing biological archives. While deep learning paradigms have gained popularity, their application to mental health genomics faces unique constraints. No prior work had resolved the specific challenges associated with sparse microarray data in schizophrenia cohorts. That uncertainty drove the need for more robust computational frameworks capable of expanding limited datasets. It was already known that generative adversarial networks provide a powerful mechanism for creating realistic synthetic information. This study addresses the persistent difficulty of training reliable models when primary patient data is scarce.

Purpose Of The Study:

The aim of this study is to introduce a modified generative adversarial network model for augmenting schizophrenia gene expression samples. Researchers sought to resolve the persistent challenge of limited biospecimen availability in psychiatric transcriptomics. This project addresses the difficulty of applying machine learning to datasets with very few patient records. The authors intended to create a framework that produces synthetic data while maintaining high statistical similarity to real-world microarray information. A key motivation was to improve the reliability of predictive models used in precision medicine. The team focused on incorporating calibration techniques to better understand model uncertainty during the generation process. By providing confidence intervals, the researchers aimed to offer a more trustworthy output for performance metrics. This effort seeks to bridge the gap between sparse clinical data and the requirements for robust computational analysis.

Main Methods:

The review approach focuses on the implementation of a modified generative adversarial network architecture for data expansion. Researchers designed the generator to utilize a bordered Gaussian distribution for input processing. To ensure output quality, the team integrated calibration techniques directly into the classification components of the model. This design choice allows for the estimation of probabilities with higher precision during the training process. The authors also incorporated confidence intervals to define the boundaries of their point estimates. By reporting expected value ranges, the study provides a transparent assessment of performance metrics. The team utilized GAN-train and GAN-test as primary quantitative benchmarks to evaluate the synthetic data. This methodological strategy ensures that the generated information remains consistent with the statistical distribution of the original microarray inputs.

Main Results:

Key findings from the literature indicate that the proposed model successfully generates synthetic data with high fidelity to original samples. The authors report that their approach effectively expands limited microarray datasets, facilitating more robust machine learning applications. Quantitative evaluations using GAN-train and GAN-test confirm that artificial gene expression profiles closely mirror the characteristics of real biological data. The researchers observed that incorporating calibration techniques significantly improves the reliability of the classification outputs. By utilizing confidence intervals, the study provides a clear range of expected values, which helps confine point estimate limitations. The results show that the model handles the challenges of sparse psychiatric samples by creating statistically representative synthetic information. This framework demonstrates that generative adversarial networks can be adapted to overcome specific barriers in transcriptomic research. The findings highlight the potential for improved model robustness when using these specialized computational techniques.

Conclusions:

The authors demonstrate that their proposed architecture effectively generates synthetic gene expression profiles for psychiatric studies. This synthesis and implications review highlights the utility of incorporating calibration techniques to improve model reliability. By utilizing confidence intervals, the researchers provide a clearer understanding of the uncertainty inherent in their generated outputs. The study confirms that artificial samples produced by this framework closely mimic the statistical properties of original microarray data. These findings suggest that augmenting small datasets can enhance the performance of downstream machine learning applications. The authors propose that their methodology offers a viable path for overcoming biospecimen shortages in rare disease research. This work underscores the importance of rigorous validation metrics when deploying generative models in clinical bioinformatics. Ultimately, the researchers show that their approach provides a trustworthy foundation for future computational investigations in mental health.

The researchers propose the MS-ACGAN model, which utilizes a generator fed by a bordered Gaussian distribution. This mechanism expands limited microarray datasets by creating synthetic gene expression profiles that statistically resemble original biological samples, thereby addressing the scarcity of psychiatric patient data.

The authors employ calibration techniques alongside confidence intervals to quantify model uncertainty. These tools provide a range of expected values for performance metrics, ensuring that the generated synthetic data remains reliable and trustworthy for subsequent analytical tasks.

Calibration is necessary because it provides insight into model uncertainty. The researchers emphasize that this step is required to improve the robustness of classifiers, allowing for more accurate probability estimations when working with limited and complex transcriptomic information.

The researchers utilize GAN-train and GAN-test as quantitative measures. These metrics serve to validate that the synthetic information generated by the framework maintains the same characteristics as the original microarray gene expression data.

The study focuses on the phenomenon of data augmentation within transcriptomics. By comparing synthetic outputs against real-world gene expression patterns, the authors measure how effectively their model overcomes the limitations of small sample sizes in schizophrenia research.

The authors propose that their framework offers a reliable solution for expanding datasets in precision medicine. They suggest that this approach mitigates the impact of biospecimen collection difficulties, particularly for conditions where patient participation is historically low.