ASD presents a significant public health challenge, with increasing prevalence worldwide. This neurodevelopmental condition exhibits a complex etiology involving both genetic and environmental factors. Although understanding the genetic underpinnings of ASD would be crucial for developing targeted interventions and therapy, the genetic predisposing factors responsible for the condition have remained largely hidden despite many recent advances. Therefore, the aim of the present study was to compare ASD-specific genetic databases to identify shared genetic components associated with autism, independent of syndromic conditions, and elucidate their biological significance by in silico analysis.

The question is legitimately raised as to what is the point of searching for target genes in the era of whole genome sequencing, and even further narrowing down the list of them based on certain considerations. However, it must not be forgotten that the vast majority of information derived from sequencing data can only be accessed after appropriate bioinformatic analysis. Furthermore, to target relevant genetic information, it is necessary to know what to look for, and the ASD-specific gene list created in the present work can provide a useful and accurate tool for this purpose.

Undoubtedly, however, estimates of the number and composition of ASD-relevant genes vary widely among research groups, used databases, and clinical sequencing panels. A recent review of a gene set associated with autism and neurodevelopmental disorders (NDD) compiled a list of 83 high-confidence and NDD candidate genes using five disease-oriented databases. Remarkably, 14 of these were found to be in common with the 20 genes identified in the present work (MECP2, WAC, GRIN2B, STXBP1, PTEN, TCF4, POGZ, DYRK1A, ADNP, AUTS2, CHD2, SYNGAP1, DDX3X, UBE3A), but it should also be noted that, unlike the present study, they included cases of NDD, developmental disorder and intellectual disability, as a broader phenotype [18]. Another approach aimed to identify autism genes in the human genome based on patterns of gene–gene interactions and topological similarity of genes in the interaction network [19]. Using 760 autism-related genes from the SFARI Gene and OMIM databases as positive controls, all human genes were prioritized for ASD susceptibility. When comparing the first 50 hits, only three were found to be in common with those found in the present work (WAC, NRXN1, UBE3A).

In addition to the database analyses, some large exome sequencing studies have also been performed to refine the list of ASD predominant genes. While only four (CHD8, PTEN, SHANK3, NRXN1) of the 53 autism-related genes identified in one study were found to be common with our gene set [20], in another work, all 20 genes were present among their 381 hits [21]. However, whereas the former study worked exclusively with samples of ASD diagnosed individuals, the latter mainly examined NDD cases. Other large-scale whole-genome or exome sequencing studies of families with children affected by ASD have primarily focused on the role of rare inherited variants in the development of the condition. Ruzzo and colleagues identified 69 genes associated with ASD risk, including 24 that passed a stringent statistical correction [22]. It is noteworthy that there is a considerable degree of overlap (11 genes) between the gene list identified by the aforementioned study and the genes selected in the present work (WAC, SHANK3, GRIN2B, POGZ, NRXN1, DYRK1A, CHD8, ADNP, CHD2, SYNGAP1, PTEN). A comparable methodology has identified 72 genes linked to ASD, which is also in substantial concordance with our curated gene list (overlapping genes: WAC, GRIN2B, STXBP1, CHD8, PTEN, SHANK3, POGZ, NRXN1, DYRK1A, ADNP, AUTS2, CHD2, SYNGAP1) [23]. In contrast, other studies that also emphasized the role of rare genetic variants demonstrated no [24] or minimal [25] overlap (SYNGAP1) with the present study. It should be noted, however, that the latter researches were conducted with relatively smaller populations of a few tens of individuals.

The comparison of gene expression levels of ASD and control samples in different tissues may also open promising perspectives. Compared to an updated list of 109 genes found to be significantly dysregulated in individuals with autism from several recent ASD expression studies, merely one (SHANK3) was found to be shared with ours [26]. A further study, which is unique within the field, compared whole genome and RNA sequencing data from postmortem dorsolateral prefrontal cortex samples of nearly two hundred individuals across prenatal and postnatal development for various neuropsychiatric conditions, including ASD [27]. Of the 97 genes identified as ASD-related, 14 exhibited overlap with the gene set identified in the present study (GABRB3, WAC, GRIN2B, STXBP1, CHD8, PTEN, TCF4, SHANK3, POGZ, NRXN1, DYRK1A, ADNP, CHD2, SYNGAP1). Furthermore, nine of these exhibited alterations in expression across the temporal developmental scale delineated in the study, with three displaying an increasing trend (WAC, STXBP1, SHANK3) and six exhibiting a decreasing trend (CHD8, TCF4, POGZ, NRXN1, DYRK1A, ADNP). However, the Human Protein Atlas data indicate that the expression of all 20 genes we delineated was observed in the human cortex, with STXBP1 exhibiting the highest expression, and only four genes (GABRB3, STXBP1, GRIN2B, and NRXN1) were specific to this tissue [28].

The partial overlap with the literature draws attention to the careful applicability of these databases, as they still contain subjective elements, both in the ranking algorithm of ASD-related genes included (which may vary significantly from database to database) and in the defining method of ASD phenotype and diagnosis [29].

The ASD phenotype is a well-defined common feature of several well-characterized genetic syndromes with quite diverse symptoms (e.g. Rett-, Fragile X- and Down syndrome, Neurofibromatosis, Tuberous sclerosis [11]). Accordingly, as in their phenotype, there is a probable overlap in their genotype as well, which was attempted to be identified in the presented work by analyzing and comparing in silico databases. The molecular biological and clinical examination of the relatively narrow set of genes and their variants thus mapped may actually bring researchers closer to elucidating the genetic predisposition of the non-syndromic cases that constitute the vast majority of ASD patients. In addition, the highly ASD related gene set selected in this work may provide guidance for the design of more targeted, population-based genetic screening tests in large samples by predicting genetic hotspots of the condition.