Feature Selection for Autism Spectrum Disorder via a Multi-Pack Cooperative Grey Wolf Optimization Framework
DOI:
https://doi.org/10.54097/67bnyv59Keywords:
Autism Spectrum Disorder (ASD), Gene Expression Feature Selection, Multi-Pack Cooperative Grey Wolf Optimization (MPC-GWO), Metaheuristic Optimization Algorithm, Biomarker Discovery.Abstract
Autism Spectrum Disorder (ASD) is a prevalent neurodevelopmental condition in childhood and adolescence. However, its heterogeneous genetic mechanisms make early diagnosis particularly challenging. Identifying robust molecular biomarkers from high-dimensional gene expression data remains a critical bottleneck. In this study, we proposed the Multi-pack cooperative grey wolf optimization (MPC-GWO) algorithm to select robust gene biomarkers for ASD using two public blood-based transcriptomic datasets (GSE25507 for training, GSE18123 for independent testing). MPC-GWO was applied to GSE25507 to identify the minimal gene subset that reliably discriminated ASD from controls. The selected biomarker genes were then evaluated on GSE18123 to test cross-dataset generalizability. After statistical preselection and MPC-GWO refinement, we identified a 24-gene signature that achieved superior classification performance on the discovery cohort (accuracy=0.808, AUC=0.823), outperforming both the all-features baseline (accuracy=0.713) and filter-based t-test selection (accuracy=0.678). Compared with LASSO (accuracy=0.732, AUC=0.756) and Random Forest (accuracy=0.678, AUC=0.737), MPC-GWO demonstrated superior classification performance. The algorithm reduced the feature set by >88% and converged within 50 iterations. On the independent cohort, the signature achieved accuracy=0.674 and AUC=0.733. Functional enrichment analysis revealed that the selected genes are strongly associated with nervous system development, Ig-like C2-type, neurodevelopmental disorders, and extracellular space, most of which have been repeatedly implicated in ASD. These results demonstrate that MPC-GWO is effective for ASD biomarker discovery. The identified gene signature showed improved classification accuracy, and its enriched biological functions provided insights into ASD-related molecular mechanisms, suggesting potential value for future ASD-related genomic research and non-invasive diagnostic exploration.
Downloads
References
[1] G. Li, Y. Cui, and J. Su, “Adaptive mechanism-based grey wolf optimizer for feature selection in high-dimensional classification,” PLOS One, vol. 20, no. 5, p. e0318903, May 2025, doi: 10.1371/journal.pone.0318903.
[2] Md. M. Rahman, O. L. Usman, R. C. Muniyandi, S. Sahran, S. Mohamed, and R. A. Razak, “A Review of Machine Learning Methods of Feature Selection and Classification for Autism Spectrum Disorder,” Brain Sci., vol. 10, no. 12, p. 949, Dec. 2020, doi: 10.3390/brainsci10120949.
[3] Q. Al-Tashi, S. J. Abdul Kadir, H. M. Rais, S. Mirjalili, and H. Alhussian, “Binary Optimization Using Hybrid Grey Wolf Optimization for Feature Selection,” IEEE Access, vol. 7, pp. 39496–39508, 2019, doi: 10.1109/ACCESS.2019.2906757.
[4] Y. Wang, Y. Yin, H. Zhao, J. Liu, C. Xu, and W. Dong, “Grey wolf optimizer with self-repulsion strategy for feature selection,” Sci. Rep., vol. 15, no. 1, p. 12807, Apr. 2025, doi: 10.1038/s41598-025-97224-8.
[5] T. R. Golub et al., “Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring,” Science, vol. 286, no. 5439, pp. 531–537, Oct. 1999, doi: 10.1126/science.286.5439.531.
[6] N. Nouri, G. Gaglia, H. Mattoo, E. De Rinaldis, and V. Savova, “GENIX enables comparative network analysis of single-cell RNA sequencing to reveal signatures of therapeutic interventions,” Cell Rep. Methods, vol. 4, no. 6, p. 100794, Jun. 2024, doi: 10.1016/j.crmeth.2024.100794.
[7] S. Mirjalili, S. M. Mirjalili, and A. Lewis, “Grey Wolf Optimizer,” Adv. Eng. Softw., vol. 69, pp. 46–61, Mar. 2014, doi: 10.1016/j.advengsoft.2013.12.007.
[8] J. Y. Khaseeb, A. Keshk, and A. Youssef, “Improved Binary Grey Wolf Optimization Approaches for Feature Selection Optimization,” Appl. Sci., vol. 15, no. 2, p. 489, Jan. 2025, doi: 10.3390/app15020489.
[9] J. Huang, X. Deng, and L. Hu, “Enhanced grey wolf optimizer with hybrid strategies for efficient feature selection in high-dimensional data,” Inf. Sci., vol. 705, p. 121958, Jul. 2025, doi: 10.1016/j.ins.2025.121958.
[10] Z. Wang, L. Shu, S. Yang, Z. Zeng, D. He, and S. Chan, “Multi-population dynamic grey wolf optimizer based on dimension learning and Laplace Mutation for global optimization,” Expert Syst. Appl., vol. 265, p. 125863, Mar. 2025, doi: 10.1016/j.eswa.2024.125863.
[11] A. Maleki, M. Roayaei, and S. Mirjalili, “Enhancing leadership-based metaheuristics using reinforcement learning: A case study in grey wolf optimizer,” Knowl.-Based Syst., vol. 330, p. 114471, Nov. 2025, doi: 10.1016/j.knosys.2025.114471.
[12] M. Banaie-Dezfouli, M. H. Nadimi-Shahraki, and Z. Beheshti, “BE-GWO: Binary extremum-based grey wolf optimizer for discrete optimization problems,” Appl. Soft Comput., vol. 146, p. 110583, Oct. 2023, doi: 10.1016/j.asoc.2023.110583.
[13] M. D. Alter et al., “Autism and Increased Paternal Age Related Changes in Global Levels of Gene Expression Regulation,” PLoS ONE, vol. 6, no. 2, p. e16715, Feb. 2011, doi: 10.1371/journal.pone.0016715.
[14] S. W. Kong et al., “Characteristics and predictive value of blood transcriptome signature in males with autism spectrum disorders,” PloS One, vol. 7, no. 12, p. e49475, 2012, doi: 10.1371/journal.pone.0049475.
[15] I. Hussain, M. Qureshi, M. Ismail, H. Iftikhar, J. Zywiołek, and J. L. López-Gonzales, “Optimal features selection in the high dimensional data based on robust technique: Application to different health database,” Heliyon, vol. 10, no. 17, p. e37241, Sep. 2024, doi: 10.1016/j.heliyon.2024.e37241.
[16] S. Berto et al., “Association between resting-state functional brain connectivity and gene expression is altered in autism spectrum disorder,” Nat. Commun., vol. 13, no. 1, p. 3328, Jun. 2022, doi: 10.1038/s41467-022-31053-5.
[17] Z. Chen et al., “Feature selection may improve deep neural networks for the bioinformatics problems,” Bioinformatics, vol. 36, no. 5, pp. 1542–1552, Mar. 2020, doi: 10.1093/bioinformatics/btz763.
[18] S. Nogueira, K. Sechidis, and G. Brown, “On the Stability of Feature Selection Algorithms,” J. Mach. Learn. Res., vol. 18, no. 174, pp. 1–54, 2018.
[19] Y. Saeys, I. Inza, and P. Larrañaga, “A review of feature selection techniques in bioinformatics,” Bioinformatics, vol. 23, no. 19, pp. 2507–2517, Oct. 2007, doi: 10.1093/bioinformatics/btm344.
[20] J. Li et al., “Feature Selection: A Data Perspective,” ACM Comput. Surv., vol. 50, no. 6, pp. 1–45, Nov. 2018, doi: 10.1145/3136625.
[21] B. Xue, M. Zhang, W. N. Browne, and X. Yao, “A Survey on Evolutionary Computation Approaches to Feature Selection,” IEEE Trans. Evol. Comput., vol. 20, no. 4, pp. 606–626, Aug. 2016, doi: 10.1109/TEVC.2015.2504420.
[22] B. H. Nguyen, B. Xue, and M. Zhang, “A survey on swarm intelligence approaches to feature selection in data mining,” Swarm Evol. Comput., vol. 54, p. 100663, May 2020, doi: 10.1016/j.swevo.2020.100663.
[23] U. M. Khaire and R. Dhanalakshmi, “Stability of feature selection algorithm: A review,” J. King Saud Univ. - Comput. Inf. Sci., vol. 34, no. 4, pp. 1060–1073, Apr. 2022, doi: 10.1016/j.jksuci.2019.06.012.
[24] Y. Wu, H.-C. Huang, and L.-X. Qin, “Making External Validation Valid for Molecular Classifier Development,” JCO Precis. Oncol., no. 5, pp. 1250–1258, Nov. 2021, doi: 10.1200/PO.21.00103.
[25] K. Li et al., “TaiChiNet: PCA-based Ying-Yang dilution of inter- and intra-BERT layers to represent anti-coronavirus peptides,” Expert Syst. Appl., vol. 282, p. 127786, Jul. 2025, doi: 10.1016/j.eswa.2025.127786.
[26] R. Zhang et al., “PepHarmony: a multi-view contrastive learning framework for integrated sequence and structure-based peptide representation,” Neural Netw., vol. 194, p. 108148, Feb. 2026, doi: 10.1016/j.neunet.2025.108148.
[27] X. Mi, B. Zou, F. Zou, and J. Hu, “Permutation-based identification of important biomarkers for complex diseases via machine learning models,” Nat. Commun., vol. 12, no. 1, p. 3008, May 2021, doi: 10.1038/s41467-021-22756-2.
[28] B. T. Sherman et al., “DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update),” Nucleic Acids Res., vol. 50, no. W1, pp. W216–W221, Jul. 2022, doi: 10.1093/nar/gkac194.
[29] H. R. Willsey, A. J. Willsey, B. Wang, and M. W. State, “Genomics, convergent neuroscience and progress in understanding autism spectrum disorder,” Nat. Rev. Neurosci., vol. 23, no. 6, pp. 323–341, Jun. 2022, doi: 10.1038/s41583-022-00576-7.
[30] J. Piñero et al., “The DisGeNET knowledge platform for disease genomics: 2019 update,” Nucleic Acids Res., p. gkz1021, Nov. 2019, doi: 10.1093/nar/gkz1021.
[31] X. Zhou et al., “Integrating de novo and inherited variants in 42,607 autism cases identifies mutations in new moderate-risk genes,” Nat. Genet., vol. 54, no. 9, pp. 1305–1319, Sep. 2022, doi: 10.1038/s41588-022-01148-2.
[32] Q. Zhang et al., “Contactin-associated protein-like 2 (CNTNAP2) mutations impair the essential α-secretase cleavages, leading to autism-like phenotypes,” Signal Transduct. Target. Ther., vol. 9, no. 1, p. 51, Mar. 2024, doi: 10.1038/s41392-024-01768-6.
[33] L. E. Rexrode et al., “Molecular profiling of the hippocampus of children with autism spectrum disorder,” Mol. Psychiatry, vol. 29, no. 7, pp. 1968–1979, Jul. 2024, doi: 10.1038/s41380-024-02441-8.
[34] H. Cheng et al., “Truncating Variants in NAA15 Are Associated with Variable Levels of Intellectual Disability, Autism Spectrum Disorder, and Congenital Anomalies,” Am. J. Hum. Genet., vol. 102, no. 5, pp. 985–994, May 2018, doi: 10.1016/j.ajhg.2018.03.004.
[35] A. R. Burmeister et al., “Maternal Herpesviridae infection during pregnancy alters midbrain dopaminergic signatures in adult offspring,” Neurobiol. Dis., vol. 169, p. 105720, Jul. 2022, doi: 10.1016/j.nbd.2022.105720.
[36] E. Diez-Itza, A. Viejo, and M. Fernández-Urquiza, “Pragmatic Profiles of Adults with Fragile X Syndrome and Williams Syndrome,” Brain Sci., vol. 12, no. 3, p. 385, Mar. 2022, doi: 10.3390/brainsci12030385.
[37] A. Khan, P. Sharma, S. Dahiya, and B. Sharma, “Plexins: Navigating through the neural regulation and brain pathology,” Neurosci. Biobehav. Rev., vol. 169, p. 105999, Feb. 2025, doi: 10.1016/j.neubiorev.2024.105999.
[38] I. M. W. Silva, J. Rosenfeld, S. A. Antoniuk, S. Raskin, and V. S. Sotomaior, “A 1.5Mb terminal deletion of 12p associated with autism spectrum disorder,” Gene, vol. 542, no. 1, pp. 83–86, May 2014, doi: 10.1016/j.gene.2014.02.058.
[39] T. Guo et al., “Histone Modifier Genes Alter Conotruncal Heart Phenotypes in 22q11.2 Deletion Syndrome,” Am. J. Hum. Genet., vol. 97, no. 6, pp. 869–877, Dec. 2015, doi: 10.1016/j.ajhg.2015.10.013.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
