ACMD Seminar: nmrPype: Utilizing Linear Algebra to Decompose NMR Signal Data

Mykah Smith
Post-baccalaureate Researcher, Delaglio Group, Institute for Bioscience and Biotechnology Research

Tuesday, Jan. 28, 2025, 3:00-4:00 PM ET (1:00-2:00 PM MT)

Abstract: Python has become a preeminent computer language for scientific computing and machine learning, and its use has been adopted in many NMR applications, notably including the nmrglue package of Helmus and Jaroniec. In order to exploit the rich set of scientific and math libraries of python in pipeline-based processing schemes, I have implemented nmrPype in collaboration with Dr. Frank Delaglio, a python analog of the widely-used nmrPipe program for spectral processing. I demonstrate the use of nmrPype to implement a convenient general-purpose linear decomposition of 1D-4D multidimensional time-domain, interferogram, and spectral data in terms of 1D, 2D, or 3D prior-knowledge basis sets established by spectral modeling. These decomposition schemes provide a way to exploit prior knowledge of signal position and lineshape in order to extract information from sparser amounts of data, and optionally without the need for complete spectral reconstruction. Applications demonstrated include extraction of signal evolutions from pseudo-3D relaxation series and pseudo-4D chemical exchange saturation transfer series, baseline correction by time-domain extrapolation, and evaluation of sampling schedules. Since the decomposition approach can extract information without the need for complete spectral reconstruction, it also provides the possibility to evaluate spectral quality during acquisition, especially in cases such as large four-dimensional spectra where it might not be practical to perform a complete reconstruction after every new increment is acquired.

Bio: Micah Smith is a post-baccalaureate researcher in the Delaglio group at the Institute for Bioscience and Biotechnology Research (IBBR), a research partnership of the University of Maryland, College Park (UMCP), the University of Maryland, Baltimore (UMB), and the National Institute of Standards and Technology (NIST). Before his post-baccalaureate research, he obtained undergraduate degrees in mathematics from Montgomery College and from University of Maryland, Baltimore County (UMBC) with a minor in Bioinformatics. At UMBC, his undergraduate research studied frameshifting in brewer’s yeast, using Python to determine potential open reading frames and resulting translations. During his time at IBBR, he has worked under Dr. Frank Delaglio and Dr. Christina Bergonzo on various computational biology projects, focusing on NMR and molecular dynamics. For Dr. Delaglio he has developed nmrPype, a companion program to industry standard NMRPipe, with feature parity and a new linear decomposition feature. In collaboration with Dr. Bergonzo, he has performed supplemental work on several molecular dynamics programs such as CPPTRAJ and MDANCE.

Host: Melinda Kleczynski

Note: This talk will be recorded to provide access to NIST staff and associates who could not be present to the time of the seminar. The recording will be made available in the Math channel on NISTube, which is accessible only on the NIST internal network. This recording could be released to the public through a Freedom of Information Act (FOIA) request. Do not discuss or visually present any sensitive (CUI/PII/BII) material. Ensure that no inappropriate material or any minors are contained within the background of any recording. (To facilitate this, we request that cameras of attendees are muted except when asking questions.)

*Safety Precaution: The hallway leading from the Courtyard to the exit closest to B-111 and B-113 will be used by contractors to move debris, machinery, and other supplies, as well as will be heavily trafficked by the contractors throughout the process. Be aware of the safety precautions posted during this time.

Note: Visitors from outside NIST must contact Meliza Lane at least 24 hours in advance.