1.1AbstractHeavy metals are toxic substances that have caused great ecological damage, can sorb with sediments, become mobile in surface waters and infiltrate the food chain. As a result, long-term exposures associated with ingestion and recreational exposures lead to environmental and human health-associated risks. The present work aims to monitor environmental and human health risk factors, using Nuclear respiratory factor 1(nrf1), transcription factor (TF) and its targets as molecular markers of heavy metal toxicity in adult zebrafish. Here, a novel Machine Learning (ML) of Molecular Stressor of Surveillance Framework (MsSF) focuses on molecular risk factors associated with heavy metal toxic stress and was designed to assist in the long-term monitoring and detection of heavy metals found in surface waters. Public datasets were mined to monitor the transcription factor activity ofnrf1and its target genes.Results shownrf1was downregulated and its responses in the hepatic system produced significant differential expression among heavy metal treated samples. Three novel Dynamic Bayesian Networks (DBN) were trained to present the optimal scenario of gene markers to discretize toxicity associated with arsenic, cadmium, and mercury. Comparative analysis in among male zebrafish described in GSE3048, GSE41622 and GSE18861 shows a total of 117 genes discovered by the DBN model. Among the toxicity responsive genes, are thenrf1molecular signatures, whereas 5 genes are common to arsenic, cadmium and mercury. Enriched stress pathways revealed apoptotic signaling pathway, transcription regulator activity and catalytic activity in all groups for male zebrafish. We observed that both gender andnrf1were associated with toxicity in regards to arsenic, cadmium and mercury. These are:chac1, dnajb11, gstp1, hspa8 and nrf1. Additionally, the toxicity interacting genes among the female zebrafish were identified.Here, during the cohort GSE30062, and GSE41623 a comparative analysis described a total of 45 toxicity marker genes, of which 10 genes are common to arsenic and cadmium. Thenrf1molecular signature genes expressed during the arsenic and cadmium exposures found that they expressed significant signaling pathway association with biological regulation, cellular process, homeostatic process, metabolic process, and response to stimulus as a result todnajb11, edm1, herpud1, hyou1, nrf1, nup133, pdia6, psmb3, psme1 and sod.This research provides evidence thatnrf1and its target genes respond as a sensitive molecular model that identified heavy metal toxicity in samples that were previously exposed to acute concentrations of arsenic, cadmium, and mercury in zebrafish. This stochastic model is of high public health and ecological relevance because traditional risk assessment methods are often limited in their ability to account for unknown concentrations, and these could potentially be further explored as surrogates for toxicity caused by heavy metal exposures.