Ecotoxicology Cubed: A roadmap for improving chemical risk assessment using high-throughput genomic tools and approaches.
Pilot project led by Prof. David Spurgeon (UKCEH)
In this Working Group, we plan to scope out a roadmap that details how different types of trait information and associated analysis tools could be organised within a digital species framework to support future in silico predictions of species responses to chemical exposure. As the first step to establishing this roadmap for trait-based assessment, we here plan to:
1) identify and catalogue the current and potential future sources of traits data for species of ecotoxicological interest; and,
2) work with experts in ecotoxicology, risk assessment and data science from UKCEH, Cardiff University and the wider community to identify strategies for data integration.
As our first step towards the use of trait data a predictive mechanistic ecotoxicological framework, we will identify and compile the source of trait data available for the top 25 most commonly studies species in ecotoxicology (as identified as from the US EPA Ecotox database). For each species, we will identify the nature of information for two major types of traits.
1. Taxonomic, physiological and anatomic traits. Traits assessed will include information on species phylogeny, body size and morphology (e.g. body surface area, volume, length, mass etc.), anatomy (exoskeleton structure, respiratory structures) respiratory structures (gills, lungs etc.), tropic level and energy budget parameters. Traits information can be drawn from existing species trait databases, past trait studies in ecotoxicology e.g. 3,4 and physiological and energetic traits database, including the AddMyPet database (www.bio.vu.nl/thb/deb/deblab/add_my_pet/).
2. Toxicokinetic and toxicodynamic traits. Limited toxicokinetics and toxicodynamics traits information are readily available. To bridge this gap, we will assess the potential for using the growing volume of genome and transcriptome data to identify 1) key metabolism pathways, enzyme system compliments and key enzyme activity measurements that determine toxicokinetics and 2) receptor and pathway components linked to known molecular initiating events that determine toxicodynamics. Until only recently, high-quality genomic resources have not been available for major ecotoxicological study organisms – the one exception being for Daphnia. This has now changed. New genomes and transcriptome are increasing available for many “non-model” species (e.g. through the Wellcome Trust Tree of Life Project). From experience, we know that all published and/or archived genomes and transcriptomes differ in quality and that this affects the nature and integrity of the analyses that can be done based on this information. Therefore, we plan to catalogue the availability of current genome resources for the “top” ecotox species. To complement the development of the genome resource catalogue, we will also scope the development of a potential bioinformatic ‘workflow’ suitable for data mining to identify orthologues of targets associated with toxicokinetic and toxicodynamic traits. This will involve working with computation and bioinformatic experts within the STFC community to identify the computational solution to data harvesting, curation, analysis and integration. Our initial assessment will be focussed on pharmacological/pesticide target, such as the TOXCAST set 5; receptors showing frequent activation in environmental samples e.g. by 6 and consensus panel adverse drug receptors 7. Comparative analyses will use sequence-based phylogenetic identification and conservation, assessment tools, such as EcoDrug8 and seqAPASS9, as well as bespoke models, such as those we have developed in our work on earthworm insecticide sensitivity assessment.