Dataset Information

Machine learning deciphers structural features of RNA duplexes measured with solution X-ray scattering.

ABSTRACT: Macromolecular structures can be determined from solution X-ray scattering. Small-angle X-ray scattering (SAXS) provides global structural information on length scales of 10s to 100s of Ångstroms, and many algorithms are available to convert SAXS data into low-resolution structural envelopes. Extension of measurements to wider scattering angles (WAXS or wide-angle X-ray scattering) can sharpen the resolution to below 10?Å, filling in structural details that can be critical for biological function. These WAXS profiles are especially challenging to interpret because of the significant contribution of solvent in addition to solute on these smaller length scales. Based on training with molecular dynamics generated models, the application of extreme gradient boosting (XGBoost) is discussed, which is a supervised machine learning (ML) approach to interpret features in solution scattering profiles. These ML methods are applied to predict key structural parameters of double-stranded ribonucleic acid (dsRNA) duplexes. Duplex conformations vary with salt and sequence and directly impact the foldability of functional RNA molecules. The strong structural periodicities in these duplexes yield scattering profiles with rich sets of features at intermediate-to-wide scattering angles. In the ML models, these profiles are treated as 1D images or features. These ML models identify specific scattering angles, or regions of scattering angles, which correspond with and successfully predict distinct structural parameters. Thus, this work demonstrates that ML strategies can integrate theoretical molecular models with experimental solution scattering data, providing a new framework for extracting highly relevant structural information from solution experiments on biological macromolecules.

SUBMITTER: Chen YL

PROVIDER: S-EPMC7467162 | biostudies-literature | 2020 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning deciphers structural features of RNA duplexes measured with solution X-ray scattering.

Chen Yen-Lin YL Pollack Lois L

IUCrJ 20200812 Pt 5

Macromolecular structures can be determined from solution X-ray scattering. Small-angle X-ray scattering (SAXS) provides global structural information on length scales of 10s to 100s of Ångstroms, and many algorithms are available to convert SAXS data into low-resolution structural envelopes. Extension of measurements to wider scattering angles (WAXS or wide-angle X-ray scattering) can sharpen the resolution to below 10 Å, filling in structural details that can be critical for biological functio ...[more]

PMID: 32939279

Similar Datasets

Project description:One of the most important questions in biological science is how a protein functions. When a protein performs its function, it undergoes regulated structural transitions. In this regard, to better understand the underlying principle of a protein function, it is desirable to monitor the dynamic evolution of the protein structure in real time. To probe fast and subtle motions of a protein in physiological conditions demands an experimental tool that is not only equipped with superb spatiotemporal resolution but also applicable to samples in solution phase. Time-resolved X-ray solution scattering (TRXSS), discussed in this Account, fits all of those requirements needed for probing the movements of proteins in aqueous solution. The technique utilizes a pump-probe scheme employing an optical pump pulse to initiate photoreactions of proteins and an X-ray probe pulse to monitor ensuing structural changes. The technical advances in ultrafast lasers and X-ray sources allow us to achieve superb temporal resolution down to femtoseconds. Because X-rays scatter off all atomic pairs in a protein, an X-ray scattering pattern provides information on the global structure of the protein with subangstrom spatial resolution. Importantly, TRXSS is readily applicable to aqueous solution samples of proteins with the aid of theoretical models and therefore is well suited for investigating structural dynamics of protein transitions in physiological conditions. In this Account, we demonstrate that TRXSS can be used to probe real-time structural dynamics of proteins in solution ranging from subtle helix movement to global conformational change. Specifically, we discuss the photoreactions of photoactive yellow protein (PYP) and homodimeric hemoglobin (HbI). For PYP, we revealed the kinetics of structural transitions among four transient intermediates comprising a photocycle and, by applying structural analysis based on ab initio shape reconstruction, showed that the signaling of PYP involves the protrusion of the N-terminus with significant increase of the overall protein size. For HbI, we elucidated the dynamics of complex allosteric transitions among transient intermediates. In particular, by applying structural refinement analysis based on rigid-body modeling, we found that the allosteric transition of HbI accompanies the rotation of quaternary structure and the contraction between two heme domains. By making use of the experimental and analysis methods presented in this Account, we envision that the TRXSS can be used to probe the structural dynamics of various proteins, allowing us to decipher the working mechanisms of their functions. Furthermore, when combined with femtosecond X-ray pulses generated from X-ray free electron lasers, TRXSS will gain access to ultrafast protein dynamics on sub-picosecond time scales.

Dataset Information

Machine learning deciphers structural features of RNA duplexes measured with solution X-ray scattering.

Publications

Machine learning deciphers structural features of RNA duplexes measured with solution X-ray scattering.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure