Quarter of a petabyte of data closer to discovering the physical principles of information storage in DNA
Scientists have conducted one of the largest studies to date using computer simulations to examine how different DNA sequences behave. The obtained data help to better understand the functioning of DNA and constitute a vast, publicly available database for further scientific research. The study results, co-authored by scientists from Gdańsk Tech, were published in the journal Nature Communications.
Scientists have conducted one of the largest studies to date using computer simulations to examine how different DNA sequences behave. The obtained data help to better understand the functioning of DNA and constitute a vast, publicly available database for further scientific research. The study results, co-authored by scientists from Gdańsk Tech, were published in the journal Nature Communications.
Fourteen groups of scientists affiliated with the ABC (Ascona B-DNA Consortium) participated in the research. This is another initiative of this research consortium, aimed at thoroughly investigating the mechanical properties of various DNA sequences using computational methods. The results of earlier stages of the long-term project allowed for the development of tools, among others, for predicting nucleosome positioning and calculating the deformation cost of any DNA sequence, while the latest results improve the accuracy of these methods. Based on them, it will be possible to create more precise models and develop AI-based tools. Moreover, each subsequent research cycle sets new standards for the analysis and visualization of DNA structural elements.
The research results were published in the journal “Nature Communications.” Four scientists affiliated with Gdańsk Tech are co-authors of the publication: Dr. Eng. Miłosz Wieczór (second author; then at IRB Barcelona, currently at the Department of Physical Chemistry, Faculty of Chemistry, Gdańsk Tech), Prof. Dr. Eng. Jacek Czub, and M.Sc. Eng. Michał Jurkowski (Department of Physical Chemistry, Faculty of Chemistry, Gdańsk Tech, Doctoral School at Gdańsk Tech), as well as Dr. Kazi Amirul Hossein (PhD in 2023 at Gdańsk Tech, at the time of publication at IRB Barcelona, currently at Schrödinger, Inc).
Tracking processes occurring in a fraction of a second
This is one of the largest studies to date in this field. The scientists collected a total of 0.25 petabytes of data (250 terabytes), enabling the description of events occurring on a time scale down to 0.1 seconds while maintaining extremely high temporal and spatial precision.
– The simulations reproduce the course of the studied process at the femtosecond scale, allowing us to observe structural changes occurring at the atomic level. However, certain DNA structural elements change over micro- and even millisecond timescales, and only months of simulation allow us to describe these relatively rare phenomena using statistical language – says Dr. Eng. Miłosz Wieczór, second author of the publication.
The data are prepared according to a unified scheme, which is important, for example, in the context of machine learning, and all simulations are available in an open database, enabling their use for further analysis.
The contribution of the Gdańsk Tech team included conducting part of the simulations within the consortium, as well as the analysis and visualization of large structural data sets using innovative deep learning tools.
– Our work on simulations of systems containing nucleic acids began about 13-14 years ago. Since then, we have had a continuous contribution to the development and improvement of new standards, an important aspect of which is the long-term collaboration with Prof. Modesto Orozco's group at the University of Barcelona and Prof. Giulia Palermo at the University of California, Los Angeles – adds Dr. Eng. Miłosz Wieczór.
