Machine Learning to Find SWCNT-Recognition DNA Sequences

Despite gains in understanding of the structure and thermodynamics of DNA/SWCNT hybrids, we have had little success in predicting which sequences are useful for separation. We have recently begun to investigate the use of machine learning techniques, trained by data sets of known sequences, to develop predictive models.  Preliminary results look very promising! (Work led by Yoona Yang and in collaboration with M. Zheng.)