Structural Insights into Hearing Loss Genetics from Polarizable Protein Repacking

Tollefson, Mallory R.; Litman, Jacob M.; Qi, Guowei; Marini, Robert J.; O’Connell, C. E.; Wipfler, M. J.; Bernabe, H.V.; Tollefson, William T.A.; Casavant, Thomas L.; Braun, Terry A.; Smith, Richard J.; Schnieders, Michael J.

doi:10.1101/556258

Cited by 3 publications

(13 citation statements)

References 45 publications

(54 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in 2015, many-body versions of both criteria were published 22 , showing provable global optimization of rotamers under many-body potentials such as AMOEBA. Beyond improving X-ray structural refinements, this many-body DEE has been applied to improving homology models for proteins involved in deafness (Chapter 2) 79 .…”

Section: ( ) ∝ Exp(− ( ) N ⁄ )mentioning

confidence: 99%

“…Here, we focus on a subset of the global optimization problem: optimal side-chain positioning. Rotamer optimization (RO) attempts to find an optimal side-chain packing (defined as minimum possible potential energy) given a fixed backbone and a discrete set of side chain placement options known as "rotamers" ( Figure 5) 22,75,[78][79]118 . Side chain optimizers have been employed in X-ray structure determination 22 , homology modeling 72 , and in ab initio prediction 119 .…”

Section: Side-chain Optimization With Pairwise Potentialsmentioning

confidence: 99%

“…For our original publication, the MB-Goldstein inequalities were truncated at ‹ terms, neglecting ô and higherorder terms 22 . For our subsequent work 79 , these were truncated at 3 terms, a decision which will be elaborated on later in the chapter. Furthermore, a slight modification makes these inequalities robust to the "neglected" higher-order energies: a buffer can be added to forbid elimination unless the criterion is satisfied with a margin for error in neglected energy terms (e.g.…”

Section: Side-chain Optimization With Many-body Potentialsmentioning

confidence: 99%

“…Tollefson found the empirically useful 10 Å box size used for the 2019 paper 79 . Several approximations were used in the course of this work.…”

Section: Box Optimization: Iterative Optimization Of Subsectionsmentioning

confidence: 99%

“…The SB algorithm builds off of our prior dual force field (DFF) algorithm, which was limited to small systems, and uses a novel sampling-based approximation to address the difficulty of large-scale IFE simulations. 22,79 , hardware 15 , and force fields 37-39, 46, 106 to generate a set of 472 homology models with repacked sidechains. From here, two main branches of progress appear: improving the optimization, and applying the produced structures to investigate hearing loss.…”

Section: The Work Of Chapter 3 Directly Addresses the Tradeoff Betweementioning

confidence: 99%

See 4 more Smart Citations

Advanced optimization and sampling techniques for biomolecules using a polarizable force field

Litman¹

View full text Add to dashboard Cite

with science and the Ph.D. process. Prof. Schnieders and the rest of the Schnieders lab provided endless support and guidance. Without his assistance, this work would not have been possible. Mallory Tollefson in particular is co-first-author on the material presented in Chapter 2, having performed most of the data collection and determined appropriate cutoff schemes. Andrew Thiel assisted in preparing the manuscript referenced in Chapter 3. Hernan Bernabe is leading the Monte Carlo Orthogonal Space Random Walk project discussed in Section 4D. Guowei Qi produced an automated script for the work described by Chapter 2, vastly reducing the human effort required and speeding up that project by months. Maya Benowitz assisted with deriving the total number of energy evaluations needed for the many-body expansion (Chapter 2). I would also like to thank the labs of Prof. Pengyu Ren, Prof. Richard Smith, Prof. Ernesto Fuentes, and Prof. Simonson. Professor Pengyu Ren and his student, Zhifeng Jing, provided data used in Chapter 3 prior to their publication, which helped us perform our study on simultaneous bookending. Prof. Richard Smith and the other members of Audiogene provided data, motivation, and critical analysis for the work in Chapter 2, which is our most directly clinically motivated work. The labs of Prof. Fuentes and Simonson assisted in the simultaneous bookending project; while we were unable to produce high-quality results on the PDZ domains iii they study, their advice and assistance was key to developing the algorithm. In addition, the Simonson lab described the native environment approximation used by the optimization work.

show abstract

Section: ( ) ∝ Exp(− ( ) N ⁄ )mentioning

confidence: 99%

Section: Side-chain Optimization With Pairwise Potentialsmentioning

confidence: 99%

Section: Side-chain Optimization With Many-body Potentialsmentioning

confidence: 99%

“…Tollefson found the empirically useful 10 Å box size used for the 2019 paper 79 . Several approximations were used in the course of this work.…”

Section: Box Optimization: Iterative Optimization Of Subsectionsmentioning

confidence: 99%

Section: The Work Of Chapter 3 Directly Addresses the Tradeoff Betweementioning

confidence: 99%

See 3 more Smart Citations

Advanced optimization and sampling techniques for biomolecules using a polarizable force field

Litman¹

View full text Add to dashboard Cite

show abstract

Protein Structure Prediction Using a Maximum Likelihood Formulation of a Recurrent Geometric Network

Tollefson

Gogal

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Only ~40% of the human proteome has structural coordinates available from experiment (i.e., X-ray crystallography, NMR spectroscopy, or cryo-EM) or homology modeling with quality templates (i.e., 30% sequence identity or greater), leaving most of the proteome structurally unsolved. Deep learning (DL) methods for predicting protein structure can help close knowledge gaps where experimental and homology models are difficult to obtain. Recent advances in these DL methods have shown promising results in expanding structural coverage to the scale of the entire human proteome, providing researchers with more complete protein structural information. Here, we improve upon an existing DL algorithm for protein structure prediction, the Recurrent Geometric Network (RGN). We first expand the training dataset to include experimental uncertainty data in the form of atomic displacement parameters, then derive a maximum likelihood loss function that incorporates this uncertainty data into model training. Compared to the original RGN, our novel maximum likelihood model improves the rate of convergence of initial model training and ultimately results in more accurate structure prediction according to the root mean square deviation (RMSD) of backbone atoms, the Global Distance Test (GDT), the Global Distance Test High Accuracy (GDT-HA), and the Template-Modeling Score (TM-Score). Our model also predicts structures with more favorable backbone torsions, which provide more accurate starting coordinates for downstream physics-based simulations. Based on these results, our maximum likelihood reformulation provides a framework for improving existing or future machine learning algorithms for protein structure prediction. The augmented dataset, data collection scripts, reformulated RGN source code, and a series of trained models are publicly available at https://github.com/SchniedersLab/likelihood-rgn.

show abstract

Nanophthalmos patient with a THR518MET mutation in MYRF, a case report

et al. 2020

View full text Add to dashboard Cite

Background Nanophthalmos has a significant genetic background and disease-causing mutations have been recently been reported in the myelin regulatory factor (MYRF) gene. We report clinical features in a patient with nanophthalmos and a Thr518Met MYRF mutation. Case presentation A three-year-old male was discovered to have nanophthalmos after first presenting to the emergency department for a frontal headache, eye pain, emesis, and lethargy. Imaging studies (CT and MRI) were negative except for increased posterior fossa cerebrospinal fluid. Subsequent examinations revealed nanophthalmos (short axial eye lengths 18.1 mm OD and 18.3 mm OS), microcornea, and a large crystalline lens. Peripheral chorioretinal pigment abnormalities were also observed. He experienced episodes of marked ocular hypertension (53 mmHg OD and 60 mmHg) likely due to intermittent angle closure precipitated by nanophthalmos. The ocular hypertension was responsive to topical medicines. Genetic analysis of known nanophthalmos genes MFRP and TMEM98 were negative, while a novel mutation, Thr518Met was detected in MYRF. The Thr518Met mutation was absent from 362 matched normal controls and was extremely rare in a large population database, allele frequency of 0.000024. The Thr518Met mutation altered a highly conserved amino acid in the MYRF protein and three of four algorithms suggested that this mutation is likely pathogenic. Finally, molecular modeling showed that the Thr518Met mutation is damaging to MYRF structure. Together these data suggest that the Thr518Met mutation causes nanophthalmos. Conclusions Nanophthalmos may present at an early age with features of angle closure glaucoma and a Thr518Met mutation in MYRF was detected in a patient with nanophthalmos. Prevalence data, homology data, mutation analysis data, and protein modeling data suggest that this variant is pathogenic and may expand the phenotypic range of syndromic nanophthalmos caused by MYRF mutations to include central nervous system abnormalities (increased posterior fossa cerebrospinal fluid).

show abstract

Structural Insights into Hearing Loss Genetics from Polarizable Protein Repacking

Cited by 3 publications

References 45 publications

Advanced optimization and sampling techniques for biomolecules using a polarizable force field

Advanced optimization and sampling techniques for biomolecules using a polarizable force field

Protein Structure Prediction Using a Maximum Likelihood Formulation of a Recurrent Geometric Network

Nanophthalmos patient with a THR518MET mutation in MYRF, a case report

Contact Info

Product

Resources

About