(6a) Nano-Optical and -Electronic Devices with Machine Learning for Biomarker Discovery and Diagnostics in Personalized Medicine
- Conference: AIChE Annual Meeting
- Year: 2019
- Proceeding: 2019 AIChE Annual Meeting
- Group: Meet the Faculty and Post-Doc Candidates Poster Session -- Sponsored by the Education Division
- Time: Sunday, November 10, 2019 - 1:00pm-3:00pm
For personalized medicine to be fully realized, progress must be made towards developing more rapid and inexpensive point-of-care devices to probe and understand the multi-omics biomarkers that signal disease susceptibility. Such devices will provide clinicians with the data they need to prescribe patients with timely, effective treatments necessary for proper care and survival. This would be a large step forward compared to cumbersome laboratory testing that can take days to weeks or diagnoses from qualitative signs and symptoms alone. New diagnostic devices have employed a range of chemical and biological detection techniques with photonic, electronic, and other advanced materials approaches for increasing sensitivity and selectivity. Despite the progress, significant improvements are still needed. Decreasing sample preparation and pretreatment steps employing complex biological extraction, amplification, and chemical labeling is critical for making devices simpler, faster, and cheaper. Additionally, emerging devices should be engineered for detecting multiple targets and multiplexed samples for broad-spectrum diagnostics.
My PhD research has focused on exploring new techniques for nucleic acid sequencing, tailored towards addressing the issues with current personalized medicine diagnostics. Sequencing not only provides a way to diagnose biomarkers on the genomic, transcriptomic, and epigenomic levels, but also serves as a platform for discovery of new biomarkers from analysis of sequenced data. While current next-generation sequencing-by-synthesis and nanopore methods have paved the way for cheaper and faster sequencing of large datasets, they still suffer from complex sample preparation requiring enzymatic amplification and chemical labeling, as well as being ensemble-based measurements with limited resolution towards single-cell studies. Through my work, I have used experimental and computational tools to develop nano-electronic and -optical spectroscopic techniques with the potential for truly single-molecule, label free nucleic acid sequencing. In two electronic methods, a scanning tunneling microscope was used for performing scanning tunneling spectroscopy and conductance measurements that probe the molecular orbitals of single nucleotides within conformationally-constrained DNA and RNA macromolecules. Robust machine learning algorithms relying on biophysical parameters were designed and implemented for identifying the nucleobases A, G, C, and T/U from their electronic signals, shown to reach accuracy levels >90%. In an optical method, surface-enhanced Raman spectroscopy measurements from nanopatterned surfaces and nanoparticle plasmonic substrates were used to measure nucleobase content in DNA k-mer fragments. A powerful bioinformatics algorithm was then designed for mapping and scoring the k-mer content measurements to a biomarker database for probabilistic identification of genes. This optical sequencing platform was used to correctly identify a Î²-lactamase antibiotic resistance gene and confirm the Pseudomonas aeruginosa pathogen of origin from <12 content measurements (<15% coverage) of the gene. The sequencing techniques developed during my PhD each have potential to facilitate rapid and inexpensive diagnostics. They are all label-free platforms requiring little to no amplification and pretreatment steps that provide broad-spectrum detection, paving the way for implementation into personalized medicine.
Future Research Directions:
As an independent investigator, I will combine my computational and experimental background in machine learning and nanoscale spectroscopic methods to build a multidisciplinary lab with the goal of discovering and detecting new targets for personalized medicine. Specifically, my lab will strive to...
- Design and apply new machine learning algorithms and bioinformatic tools for identifying novel disease biomarkers from next-generation sequencing data.
- Engineer new nano-optical and -electronic devices to detect disease biomarkers at higher resolution, pushing the limits of single-molecule sensing.
âQuantum Techniques for Single-Molecule Nucleic Acid Sequencingâ Under the supervision of Prashant Nagpal and Anushree Chatterjee, Department of Chemical and Biological Engineering, University of Colorado Boulder
National Science Foundation Graduate Research Fellowship (NSF GRFP)
Graduate Assistantship in Areas of National Need (GAANN), University of Colorado Boulder â Biomaterials
Ph.D., Chemical Engineering, University of Colorado Boulder, Expected May 2020
M.S., Chemical Engineering, University of Colorado Boulder, 2017
B.S., Chemical Engineering, University of Nebraska-Lincoln, 2014
During my PhD, I have had the opportunity to TA for a general chemistry course for first- and second-year multidisciplinary engineering students, and a tissue engineering and medical devices elective course for upper-class chemical engineers. As an advanced TA for these classes, I was involved in grading assignments, reports, and exams along with leading recitation sections, office hours, and giving guest lectures. These experiences have shaped my teaching values and motivations. I believe that students can benefit most when guided and taught a general process for problem solving (i.e., approaches applicable to many problems), and how to ask the right questions when approaching a new problem with unknown solution. This idea is very applicable to the fundamental chemical engineering curriculum in reactor design, thermodynamics, kinetics, and fluid mechanics. I think it is detrimental for students to follow specific solution procedures, where they become comfortable with redundantly and blindly applying the same set of steps for routine, elementary solutions. Students should be prepared for the unexpected, and often non-trivial problems encountered in the real world, whether in a laboratory R&D setting or industry.
Future Teaching Directions:
As a faculty member, I will strive to establish new elective courses within the fundamental chemical engineering curriculum. I will use my background in machine learning and nanobiotechnology to create courses where students can become familiar with state-of-the-art technologies and their applications. I believe that more hands-on experience, specifically with applying computer science aspects to the traditional chemical engineering skillset, is vital for students to excel in today's workforce where computational data analysis is becoming a required expertise. I would design a course where students can become familiar applying an array of machine learning algorithms towards analysis of real datasets, such as those from next-generation sequencing. Additionally, I will require students to read and critically analyze reports from scientific journals. As a graduate student, I have already begun developing such courses. I am currently designing a hands-on set of experiments with guided data analysis protocols, aimed at chemical and biological engineering senior design labs. The lab module will incorporate experiments employing the optical sequencing platform developed in my PhD work. The hope is that this will not only highlight the transformative research taking place within the department, but inspire students to pursue research careers or further education.
Service and Outreach:
In addition to teaching, I have taken part in numerous service and outreach efforts in my graduate work. I have helped organize and lead demonstrations for a science and engineering discovery day for middle school students at my university, with activities ranging from exploring Newtonâs laws of motion to polymer chemistry. As part of a multi-department cohort, I helped design a four-week nanoscience lecture and experimental demonstration series aimed at motivating high school seniors to purse STEM careers. Students were introduced to general ideas in molecular forces, self-assembly chemistry, and the everyday uses of nanotechnology from filtration membranes to surface coatings. I have also volunteered with the Junior Achievement Program, where I led elementary students in activities teaching the importance of economics within the community while providing specific input from an engineer and scientist perspective. As a faculty member, I will establish new outreach opportunities within my research group and department to motivate and excite the next generation of engineers and scientists.
 Korshoj, L. E.; Nagpal, P. Diagnostic Optical Sequencing. bioRxiv preprint: 10.1101/685792.
 Korshoj, L. E.; Nagpal, P. BOCS: DNA k-mer Content and Scoring for Rapid Genetic Biomarker Identification at Low Coverage. Comput. Biol. Med. 2019, 110, 196â206.
 Abel, Jr., G. R.; Korshoj, L. E.; Otoupal, P. B.; Khan, S.; Chatterjee, A.; Nagpal, P. Nucleotide and Structural Label Identification in Single RNA Molecules with Quantum Tunneling Spectroscopy. Chem. Sci. 2019, 10 (4), 1052â1063.
 Sagar, D. M.; Korshoj, L. E.; Hanson, K. B.; Chowdhury, P. P.; Otoupal, P. B.; Chatterjee, A.; Nagpal, P. High-Throughput Block Optical DNA Sequence Identification. Small 2018, 14 (4), 1703165.
 Korshoj, L. E.; Afsari, S.; Chatterjee, A.; Nagpal, P. Conformational Smear Characterization and Binning of Single-Molecule Conductance Measurements for Enhanced Molecular Recognition. J. Am. Chem. Soc. 2017, 139 (43), 15420â15428.
 Afsari, S.; Korshoj, L. E.; Abel, G. R.; Khan, S.; Chatterjee, A.; Nagpal, P. Quantum Point Contact Single-Nucleotide Conductance for DNA and RNA Sequence Identification. ACS Nano 2017, 11 (11), 11169â11181.
 Korshoj, L. E.; Afsari, S.; Khan, S.; Chatterjee, A.; Nagpal, P. Single Nucleobase Identification Using Biophysical Signatures from Nanoelectronic Quantum Tunneling. Small 2017, 13 (11), 1603033.
 Korshoj, L. E.; Zaitouna, A. J.; Lai, R. Y. Methylene Blue-Mediated Electrocatalytic Detection of Hexavalent Chromium. Anal. Chem. 2015, 87 (5), 2560â2564.