The key terms ‘Child abuse’, ‘Child sexual abuse’, ‘Child maltreatment’, ‘Child neglect’, ‘Child physical abuse’ and ‘Child sexual abuse’ were entered into the MEDLINE and PubMed databases without date restrictions. English-language publications related to these terms were searched, transformed from the PubMed online database (https://pubmed.ncbi.nlm.nih.gov/) and merged (downloaded and merged) into Medline format. PubMed is a search engine for life sciences and biomedical articles as it contains abstracts and references, but often also full articles with open access. It is part of the NCBI (National Center for Biotechnology Information) (12).

In this step, key terms related to abuse, maltreatment and neglect were extracted. The process was carried out in the bioinformatics environment of MATLAB Bioinformatics Toolbox using regular expression and the corresponding PUBMED ID, during which the data was evaluated in order to extract and store the results. Using bioinformatics techniques, publications from different datasets were filtered to remove publications that appear more than once. This was performed in order to create the complete set of unique publications.

SNPs associated with child abuse were extracted and annotated from various genomic databases. The identified SNPs were located in genes previously implicated in stress response and neurodevelopment. Additionally, non-coding regions, including long intergenic non-coding (LINC) RNAs and provisional gene identifiers (LOC), were analyzed.

LINC RNA genes represent a category of long non-coding RNAs that do not overlap with protein-coding genes and are involved in gene regulation. LOC genes serve as provisional gene identifiers assigned in genomic databases when their precise function or official gene symbol has yet to be determined.

For annotation, databases such as dbSNP (https://www.ncbi.nlm.nih.gov/snp/), GWAS Catalog (https://www.ebi.ac.uk/gwas/) and Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.genome.jp/kegg/) were utilized to classify the SNPs based on their genomic location, functional impact, and association with biological pathways.

In this step, the annotated gene targets were scored through semantic analysis in order to identify and isolate the most direct and likely targets associated with child abuse. Thus, a scoring function was created in which we can differentiate the ‘related SNPs’ as well as the ‘strongly related SNPs’, which were ranked and rated in terms of the frequency of occurrence of polymorphisms in the articles studied and the frequency of occurrence of polymorphisms based on each genetic target.

The score genetic targets were as follows:

where i is the number of corresponding SNPs per genetic target and n is the total number of the SNPs per genetic target. Scoring function was calculated as follows: Related_Genetic Targets: Score <3, strongly related genetic target: score ≥3.

All SNPs were identified and classified into these two main categories, strongly related genetic targets and related genetic targets based on their correlation rate with the particular terms studied.

From group A ‘Strongly related Genetic Target’ of genetic targets with high correlation, their related SNPs were analyzed in order to collect all disease ontologies. A bioinformatics algorithm was then used to identify the disease ontologies that occur with higher frequency. This precise result of the most frequently appearing key terms was captured and visualized as a word cloud in Fig. 1.

This word cloud visualizes the most frequently occurring terms related to child abuse in the dataset. Larger words indicate higher frequency within the dataset, highlighting the central themes related to child maltreatment and neglect.

In the same manner, using genetic targets, genes, pseudogenes, primates, characterized regions of dark DNA matter, such as long non coding RNAs and microRNAs (miRNAs/miRs), categories of biological pathways were discovered using the platform Reactome, in order to extract beneficial information regarding the major biological pathways of the key genetic targets.

The concept of ‘adversity genes’, as described by Levine et al (13) in 2017, is relevant to this analysis. Adversity genes refer to genetic elements that exhibit differential expression in response to environmental stressors, including childhood maltreatment. These genes, which include FKBP prolyl isomerase 5 (FKBP5), corticotropin releasing hormone receptor 1 (CRHR1), catechol-O-methyltransferase (COMT) and several non-coding RNA regions, play a pivotal role in the stress response of the body and the development of stress-related disorders (13). Their involvement in neuroendocrine signaling, epigenetic modulation, and immune system regulation further supports the hypothesis that adverse childhood experiences leave a lasting imprint at the genetic and epigenetic levels.

All suspected SNPs potentially linked to epigenetics were studied through publications in order to understand the contribution of epigenetics to this subject. To obtain this, a search for specific polymorphisms corresponding to the genetic targets was performed and re-evaluated for their epigenetic contribution to this subject.

The suspected epigenetic polymorphisms were defined as polymorphisms corresponding to non-coding regions such as long non-coding RNAs (lncRNAs), LINCs RNAs, miRs and LOCs. The following terms were used:

LINC RNAs. A type of lncRNA that does not overlap protein-coding genes and plays a role in gene regulation, chromatin remodeling, and cellular processes.

LOC (locus). A placeholder name for genes or genomic regions that have been identified but are not yet fully characterized or named. These regions may contain functional elements, including non-coding RNAs or pseudogenes.

GWAS. A research approach that involves scanning complete sets of DNA (genomes) from multiple individuals to identify genetic variants associated with specific traits or diseases.

Epigenetics. The study of changes in gene expression that do not involve alterations in the DNA sequence, but are influenced by environmental and developmental factors, such as DNA methylation and histone modifications.

DNA methylation. A biochemical process that adds a methyl group to DNA, often silencing gene expression and playing a key role in regulating development and disease.

Histone modification. Chemical changes to histone proteins that affect the structure of chromatin and influence gene expression.

SNP. A variation in a single nucleotide at a specific position in the genome that may be associated with genetic predispositions to diseases or traits.

HPA axis. A major neuroendocrine system that regulates stress responses, metabolism, immune function, and mood through the release of hormones such as cortisol.

A systematic data mining and semantic analysis approach was employed to identify genes, variants and SNPs associated with child abuse, maltreatment and neglect. A total of 201,121 publications containing key terms, such as ‘Child abuse’, ‘Child sexual abuse’, ‘Child maltreatment’, ‘Child neglect’ and ‘Child physical abuse’ were retrieved from the MEDLINE and PubMed databases. The extracted data were filtered and processed to generate a dataset of genetic factors relevant to child abuse. The key words used in this analysis are closely related to the research subject and represent the most frequently occurring terms in the relevant literature (Table I).

A visual representation of the identified key terms is presented in Fig. 1, which illustrates the frequency of key terms related to child abuse within the dataset. This visualization highlights the most commonly occurring terminology in the literature, providing insight into the breadth and focus of existing research. Additionally, a word cloud was generated to visually represent the most prominent terms extracted from the dataset (Fig. 2).

A total of 529 abuse-related keywords were identified within the dataset, including ‘Child sexual abuse’, ‘Child emotional abuse’ and ‘Child neglect’. Semantic analysis was applied to refine the dataset, ensuring the inclusion of only the most relevant terms. The frequency of these key terms was analyzed, and the most frequently occurring words were identified (Table II). The table provides an overview of the most frequently occurring key terms within the dataset, demonstrating the breadth of terminology used in the literature related to child abuse.

A visual representation of the identified key terms was created in the form of a word cloud (Fig. 2). The figure illustrates the most common key terms extracted from the dataset using semantic analysis. Words that appear more frequently in the relevant literature are depicted in larger fonts, illustrating the conceptual emphasis of the dataset.

A total of 209 SNPs and 104 genetic targets, including genes, pseudogenes and transcription factors, were identified and extracted from online databases. The SNPs associated with child abuse were annotated using MATLAB algorithms, incorporating genomic information from dbSNP, GWAS Catalog and KEGG. The genes most frequently associated with child abuse were classified according to their functional role in stress response and neurodevelopment. The identified genetic targets and SNPs related to child abuse are listed in Table III. This table compiles the key genetic targets identified in association with child abuse. It includes genes, transcription factors, pseudogenes and non-coding RNAs, highlighting their role in neurodevelopmental and stress-related biological processes.

The distribution of genetic targets and evidence found in the present study is visually presented in Fig. 3. This figure presents a graphical representation of the most frequently occurring genetic targets and non-coding regions associated with child abuse. The size of each genetic term corresponds to its relative frequency within the dataset, highlighting key genes, such as FKBP5, CRHR1 and COMT.

A semantic analysis of the extracted SNPs revealed a genomic map of child abuse-related genetic targets. The genes, FKBP5, CRHR1, LINC02210-CRHR1 and COMT, were identified as the most frequently occurring targets. These genes are associated with stress response regulation, emotional resilience and neuroendocrine signaling pathways.

FKBP5. The FKBP5 gene is a part of the immunophilin proteins, which play a role in immune regulation and functions as a co-chaperone in glucocorticoid receptor activity in response to stressors, making it one of the most frequently encountered genes in studies of people who have undergone stress, and in particular, in children who have been abused (14). In addition to epigenetic modifications and other environmental factors, it appears that FKBP5 may modulate GR susceptibility by delaying or reducing its transcriptional activity (15).

CHRH1. The CRHR1 gene has been extensively studied due to its implication for sensitized reactivity in stressful conditions (16). Through these findings, a significant interaction of CHRH1 with childhood abuse and trauma and history of suicide attempts emerges. In the similar direction, in another study on 235 HPA axis SNPs, a trend of the rs2664008 polymorphism of the CRHR1 gene, early childhood abuse and suicide attempts in bipolar patients was indicated (17).

COMT. Variations in some genes, including the COMT gene, are known to be associated with susceptibility to stress and some mental disorders. The association between stressful events and genes is known to activate the mechanism of depression development (5).

Variations in these genes are associated with stress sensitivity and depressive cognitive biases. The interaction between genes and stressful events in childhood is considered to be a mechanism that plays a role in the development of depression and therefore helps to proactively identify symptoms of depression or other diseases through genetic susceptibility (5).

The analysis of genetic targets revealed that the most common disorders associated with child abuse include neoplasms, depressive disorders, schizophrenia, neurodegenerative diseases and autoimmune conditions. These disorders share key biological pathways involved in the stress response, neurodevelopment and immune function. The most frequently occurring disease ontologies linked to child abuse are summarized in Table IV. The table categorizes the primary disease ontologies identified in the dataset, emphasizing conditions frequently associated with child abuse, such as neuropsychiatric disorders, metabolic syndromes, and immune-related pathologies.

A visual representation of the key disease associations with child abuse is provided in Fig. 4, illustrating the most prominent biological ontologies derived from the analysis. A word cloud displaying the most frequently identified biological ontologies related to child abuse is presented in Fig. 4. Terms related to neurodevelopmental processes, stress response and immune system regulation appear prominently, emphasizing their significance in the dataset.

Further overrepresentation analysis of genetic targets demonstrated marked enrichment in biological pathways related to HPA axis regulation, dopaminergic signaling and synaptic plasticity. These pathways are crucial in the adaptation of the body to stress and have been implicated in mental health disorders commonly observed in individuals with a history of child abuse. The identified pathways are summarized in Table V. This table details the biological pathways enriched in the dataset, focusing on molecular mechanisms linked to stress response, neurodevelopment and immune regulation. The pathways were identified using overrepresentation analysis.

A graphical representation of these enriched pathways is provided in Fig. 5, highlighting the statistical significance of biological processes associated with child abuse-related genetic targets. This heatmap visually represents the statistical significance (P-values) of the biological pathways linked to the most frequently occurring genes in child abuse, grouped by functional categories such as immune signaling, neurodevelopment, and stress-related mechanisms. Different colors correspond to varying levels of statistical significance, with darker shades representing lower P-values and greater relevance. The exact P-values are presented numerically within each cell of the heatmap, and statistical significance is shown using the -log₁₀(P-value) scale, allowing for both visual and quantitative interpretation.

Table VI provides a detailed breakdown of the most frequently identified biological pathways and their associated genetic targets. It highlights key signaling mechanisms implicated in the biological response to childhood adversity. The most frequently occurring biological pathways associated with adversity genes are detailed in Table VI.

An analysis of epigenetic targets identified key non-coding RNAs (LINC, LOC and miR) that may mediate the effects of child abuse. miR-195, implicated in breast cancer, was found to be epigenetically regulated, while LOC105369506 and LOC100287329 were associated with stress-related epigenetic modifications (18). These findings suggest that epigenetic alterations in non-coding genomic regions may contribute to the long-term consequences of childhood adversity. The epigenetically significant genetic targets related to child abuse are compiled in Table VII.

miR-195 is a type of miRNA that has been implicated in breast cancer. miRNAs are small RNA molecules that regulate gene expression by binding to messenger RNA (mRNA) molecules and preventing their translation into proteins. Research has demonstrated that the expression of miR-195 is downregulated, or less active, in breast cancer cells compared to normal breast tissue. This suggests that it may play a role in the development or progression of breast cancer. The expression of miR-4761 is regulated by histone modifications in breast cancer cells, suggesting that epigenetic changes may contribute to its dysregulation in breast cancer (18).

LOC105369506 is a gene that has been studied in relation to various behaviors and conditions, and research has suggested that certain epigenetic modifications can play a role in the development of antisocial behavior and the effects of child abuse. For example, studies have found that individuals who have experienced abuse or neglect during childhood may have differences in epigenetic marks on genes related to stress response and emotional regulation, which may increase their risk for antisocial behavior later in life (19-21).

LOC100287329 is a gene that is also known as miR-548AA2. It is a miRNA gene located on chromosome 14 in humans. Abuse, whether physical, emotional, or sexual, can have a significant impact on the health of an individual. There is evidence to suggest that exposure to stress and trauma can lead to epigenetic changes that may contribute to the development of MS and other diseases (22).