NIST only participates in the February and August reviews.
The aim of our group to use artificial intelligence to understand chemistry. Our current research objective is to design and train deep neural networks to understand the structure of chemicals and chemical reactions in a manner similar to how humans learn chemistry, but over sets of data that are much larger than can be understood by a single human. The resulting networks would then be used to comprehensively predict chemical reactions and properties, and potentially understand particular reaction classes at a deeper level than current scientific knowledge.
The resulting models would be used to predict reactions of interest to chemists and biologists. A particular application would be to generate theoretical spectra for molecules not yet included in the NIST mass spectral libraries, used by many thousands of scientists worldwide.
NIST is a center for measurement excellence, and, as a result, has available for this research large sets of high quality data on chemical structures and chemical reactions essential for deep learning. The spectra in the NIST libraries and their corresponding retention indices contain the high quality measurements of million of reactions. NIST also provides excellent computational support, including substantial GPU compute clusters and access to the cloud. Our group maintains python libraries and computational services that significantly simplify the use of our data in machine learning.
The successful applicant would have an opportunity to interact with the large community of chemists, physicists and computer scientists at NIST. The artificial intelligence interest group at NIST consists of over 100 scientists who interact regularly via meetings, seminars and informal gatherings.
It is not a requirement to be a chemist to apply for this position. Computer scientists, physicists, mathematicians and scientists from related disciplines will be well qualified.
Artificial Intelligence; Machine Learning; Deep Learning; Chemistry; Cheminformatics; Data science; Informatics; Mass spectrometry