How safe are wild-caught salmons exposed to various industrial chemicals? First ever in silico models for salmon toxicity data gaps filling

Siyun Yang, Supratik Kar

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Salmons are crucial to ecosystems and economic activities like commercial fishing and aquaculture, while also serving as an important source of nutrients, underscoring their ecological significance and the need for sustainable management. To better understand the toxicity and biological interactions between the salmon and industrial chemicals in the aquatic environment, we utilized the ToxValDB database to develop first ever computational toxicity models for six salmon subspecies (covering Atlantic and Pacific salmon) across two genera, employing Quantitative Structure-Activity Relationship (QSAR) and quantitative Read-Across Structure-Activity Relationship (q-RASAR) methods. For three smaller datasets (Oncorhynchus nerka, Oncorhynchus keta, and Oncorhynchus gorbuscha), we created mathematical models using the entire datasets where QSAR models demonstrated superior statistical quality compared to q-RASAR. Conversely, the three larger datasets (Oncorhynchus kisutch, Oncorhynchus tshawytscha, and Salmon salar) were divided into training and test sets, the q-RASAR models yielded better results compared to QSAR models. Mechanistic interpretations of these models revealed that descriptors such as Burden eigenvalues (BCUT), autocorrelation of topological structure (ATSC), and molecular polarizability were significant predictors of toxicity. For instance, higher polarizability and certain topological features were associated with increased toxicity as per the developed models. Statistically superior models for each subspecies were used to predict the aquatic toxicity of 1085 untested organic chemicals for toxicity data gap filling and risk assessment considering the applicability domain (AD). These insights are pivotal for designing safer chemicals and emphasize the need for sustainable management of salmon populations.

Original languageEnglish
Article number135401
JournalJournal of Hazardous Materials
Volume477
DOIs
StatePublished - 15 Sep 2024

Keywords

  • LC
  • Q-RASAR
  • QSAR
  • Risk assessment
  • Salmon
  • USEPA

Fingerprint

Dive into the research topics of 'How safe are wild-caught salmons exposed to various industrial chemicals? First ever in silico models for salmon toxicity data gaps filling'. Together they form a unique fingerprint.

Cite this