TRANSCRIPTOMIC BLOOD META-ANALYSIS IN AUTISM EXPRESSION PUBLIC DATA

Hudson Pereira; Eduardo Fukutani Rocha; Artur Queiroz; Alexandre Rossi Paschoal

All Papers

Submission

Paper Title

TRANSCRIPTOMIC BLOOD META-ANALYSIS IN AUTISM EXPRESSION PUBLIC DATA

Authors

Hudson Pereira
Eduardo Fukutani Rocha
Artur Queiroz
Alexandre Rossi Paschoal

Modality

Xpress presentation

Subject area

Systems Biology and Modeling

Publishing Date

26/04/2022

Country of Publishing

Brasil

Language of Publishing

Inglês

Paper Page

https://www.even3.com.br/anais/xmeetingxp2021/418938-transcriptomic-blood-meta-analysis-in-autism-expression-public-data

ISBN

978-65-5941-645-5

Keywords

Keywords: Autism Spectrum Disorder, Meta-Analysis, Machine Learning

Summary

Autism Spectrum Disorder (ASD) syndrome is characterized by social interaction difficulties, qualitative deviations in communication, and repetitive behaviors. This syndrome is also defined as loss of contact to reality, caused by impossibility or great difficulty in interpersonal communication. ASD is classified into three severity degrees: mild, moderate, and severe. The early diagnosis of a child with autism is essential for effective treatment. In case, bioinformatics and high throughput could be powerful too to bring evidence of gene regulatory mechanisms on ASD. In particular, transcriptomic analyses provide the information to better address the intermediate steps between genes and their biological interaction. In this work, we performed a meta-analysis on publicly available gene expression data from ASD-associated studies. First, we were mining the gene expression database (SRA) for RNA-Seq data and ASD. We found a total of 88 datasets, where 6528 samples were retrieved. Next, we manually check and filter all these results to obtain only three blood datasets (GSE129808, GSE67528, GSE61476) for further gene expression analysis. We performed an analysis to identify the differentially expressed genes in the ASD condition. We built an R script to do automatic analysis. In summary, we used (i) Fastq-dump for sample download, (ii) Trimmomatic for quality control, (iii) Salmon was mapped to reads and applied to data quantification, and finally, we used EdgeR DESq2 for expression analysis differential. All the datasets were normalized. Moreover, machine-learning approaches were employed to identify the minimal gene set to classify the normal and autism samples. In the end, we found 29373 genes. In the isolated dataset of GSE61476, we found a correlation variation of 59.9%, GSE67528 27.7%, and finally GSE129808 with 51.4%. Finally, enrichment based on the GO gene ontology was applied. This job applied data mining techniques to further elucidate the gene's importance and investigate potentially hidden connections within the gene expression datasets.

Title of the Event: X-Meeting XPerience 2021
Title of the Proceedings of the event: X-Meeting presentations
Name of the Publisher: Even3
Means of Dissemination: Meio Digital

How to cite

PEREIRA, Hudson et al.. TRANSCRIPTOMIC BLOOD META-ANALYSIS IN AUTISM EXPRESSION PUBLIC DATA.. In: X-Meeting presentations. Anais...São Paulo(SP) AB3C, 2021. Available in: https//www.even3.com.br/anais/xmeetingxp2021/418938-TRANSCRIPTOMIC-BLOOD-META-ANALYSIS-IN-AUTISM-EXPRESSION-PUBLIC-DATA. Access in: 22/03/2025