| 1. |
- Carlsson, Henrik, 1987-, et al.
(författare)
-
Combining the targeted and untargeted screening of environmental contaminants reveals associations between PFAS exposure and vitamin D metabolism in human plasma.
- 2023
-
Ingår i: Environmental Science. - : Royal Society of Chemistry. - 2050-7887 .- 2050-7895. ; 25:6, s. 1116-1130
-
Tidskriftsartikel (refereegranskat)abstract
- We have developed, validated, and applied a method for the targeted and untargeted screening of environmental contaminants in human plasma using liquid chromatography high-resolution mass spectrometry (LC-HRMS). The method was optimized for several classes of environmental contaminants, including PFASs, OH-PCBs, HBCDs, and bisphenols. One-hundred plasma samples from blood donors (19-75 years, men n = 50, women n = 50, from Uppsala, Sweden) were analyzed. Nineteen targeted compounds were detected across the samples, with 18 being PFASs and the 19th being OH-PCB (4-OH-PCB-187). Ten compounds were positively associated with age (in order of increasing p-values: PFNA, PFOS, PFDA, 4-OH-PCB-187, FOSA, PFUdA, L-PFHpS, PFTrDA, PFDoA, and PFHpA; p-values ranging from 2.5 × 10-5 to 4.67 × 10-2). Three compounds were associated with sex (in order of increasing p-values: L-PFHpS, PFOS, and PFNA; p-values ranging from 1.71 × 10-2 to 3.88 × 10-2), all with higher concentrations in male subjects compared with female subjects. Strong correlations (0.56-0.93) were observed between long-chain PFAS compounds (PFNA, PFOS, PFDA, PFUdA, PFDoA, and PFTrDA). In the non-targeted data analysis, fourteen unknown features correlating with known PFASs were found (correlation coefficients 0.48-0.99). Five endogenous compounds were identified from these features, all correlating strongly with PFHxS (correlation coefficients 0.59-0.71). Three of the identified compounds were vitamin D3 metabolites, and two were diglyceride lipids (DG 24:6;O). The results demonstrate the potential of combining targeted and untargeted approaches to increase the coverage of compounds detected with a single method. This methodology is well suited for exposomics to detect previously unknown associations between environmental contaminants and endogenous compounds that may be important for human health.
|
|
| 2. |
- Sreenivasan, Akshai P., et al.
(författare)
-
Predicting protein network topology clusters from chemical structure using deep learning
- 2022
-
Ingår i: Journal of Cheminformatics. - : BioMed Central. - 1758-2946. ; 14:1
-
Tidskriftsartikel (refereegranskat)abstract
- Comparing chemical structures to infer protein targets and functions is a common approach, but basing comparisons on chemical similarity alone can be misleading. Here we present a methodology for predicting target protein clusters using deep neural networks. The model is trained on clusters of compounds based on similarities calculated from combined compound-protein and protein-protein interaction data using a network topology approach. We compare several deep learning architectures including both convolutional and recurrent neural networks. The best performing method, the recurrent neural network architecture MolPMoFiT, achieved an F1 score approaching 0.9 on a held-out test set of 8907 compounds. In addition, in-depth analysis on a set of eleven well-studied chemical compounds with known functions showed that predictions were justifiable for all but one of the chemicals. Four of the compounds, similar in their molecular structure but with dissimilarities in their function, revealed advantages of our method compared to using chemical similarity.
|
|
