To achieve this objective, I used the NWChem software, which is a high performance computational chemistry software developed by PNNL scientists. Through a collaboration set-up by Dr. Dayle Smith (previously in Core C), I spent two months learning how to use NWChem under the supervision of Dr. Kurt Glaesemann.
The Environmental and Molecular Science Laboratory (EMSL) is a national user facility that houses multiple instruments and laboratories. My office was located inside this building at PNNL in Richland, WA.
Using this approach, my goal was to be able to predict which OHPAHs and OPAHs are likely to form in the environment based on their thermodynamic properties, specifically the reaction Gibbs free energy. There were three main areas related to Project 5’s focus where this predictive capability will be helpful. First, the results could assist in explaining why toxicity in remediated soils increased, even after PAHs’ concentrations went down (e.g. Chibwe et al., 2015). Secondly, prior data of OHPAHs found in human urine, (e.g. Motorykin et al., 2015) can be compared with computational results to see if I can formulate a prediction of which OHPAHs are likely to form in human urine. Finally, continuing on prior work that has predicted the formation of NPAHs in ambient air (e.g. Jariyasopit et al., 2013), I could then apply similar approach, but for OHPAHs and OPAHs.
The learning curve during my externship was quite steep. Although I was able to understand how to use NWChem, I also learned that a one-size-fits-all approach was not possible and I would need to tailor my modeling approach to successfully predict formation of OHPAHs and OPAHs. It was during this time that I found out how valuable it was for me to be able to spend time at PNNL. Being in a facility where there were experts in almost every imaginable field, I was able to talk to many experts about issues that I faced. These conversations led me to the field of chemometrics which helped me tailor the computational chemistry approach accordingly. One of the online programs that I found to be useful was XenoSite, which can predict CYP450 inhibition sites on a given compound. This program can potentially be useful when narrowing down potential OHPAHs that might form through bioactivation.
In addition, the myriad instruments and facilities that are available at PNNL have also assisted me greatly during my internship. For example, the NWChem software that I used was connected to the supercomputing facility, which helped speed up the calculation, resulting in faster computational time. Another perk of being in Richland, was that I managed to tour the Hanford B Reactor—coincidentally a couple of days before the anniversary of the atomic bomb being dropped in Japan.
All in all, the externship challenged me to get me out of my comfort zone, but also rewarded me with a new skill and unique
Constance, my “instrument” over the summer. Constance is one of several supercomputers available at PNNL.
experience. In a way, the internship at PNNL served as a preview of what may come once I am finished with my Ph.D. More importantly, I found the networking opportunity and exposure to a possible career path while at PNNL to be invaluable. Currently, I am excited to combine the in silico approach that I learned at PNNL with the analytical chemistry approach at the Simonich lab into my research projects. The chemical analysis component will verify how accurate the prediction capabilities are. If this approach is proven to be reliable, I hope that this perspective can offer a different insight in predicting the formation of OHPAHs and OPAHs.