skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Mutations that Cause Human Disease: A Computational/Experimental Approach

Abstract

International genome sequencing projects have produced billions of nucleotides (letters) of DNA sequence data, including the complete genome sequences of 74 organisms. These genome sequences have created many new scientific opportunities, including the ability to identify sequence variations among individuals within a species. These genetic differences, which are known as single nucleotide polymorphisms (SNPs), are particularly important in understanding the genetic basis for disease susceptibility. Since the report of the complete human genome sequence, over two million human SNPs have been identified, including a large-scale comparison of an entire chromosome from twenty individuals. Of the protein coding SNPs (cSNPs), approximately half leads to a single amino acid change in the encoded protein (non-synonymous coding SNPs). Most of these changes are functionally silent, while the remainder negatively impact the protein and sometimes cause human disease. To date, over 550 SNPs have been found to cause single locus (monogenic) diseases and many others have been associated with polygenic diseases. SNPs have been linked to specific human diseases, including late-onset Parkinson disease, autism, rheumatoid arthritis and cancer. The ability to predict accurately the effects of these SNPs on protein function would represent a major advance toward understanding these diseases. To date several attemptsmore » have been made toward predicting the effects of such mutations. The most successful of these is a computational approach called ''Sorting Intolerant From Tolerant'' (SIFT). This method uses sequence conservation among many similar proteins to predict which residues in a protein are functionally important. However, this method suffers from several limitations. First, a query sequence must have a sufficient number of relatives to infer sequence conservation. Second, this method does not make use of or provide any information on protein structure, which can be used to understand how an amino acid change affects the protein. The experimental methods that provide the most detailed structural information on proteins are X-ray crystallography and NMR spectroscopy. However, these methods are labor intensive and currently cannot be carried out on a genomic scale. Nonetheless, Structural Genomics projects are being pursued by more than a dozen groups and consortia worldwide and as a result the number of experimentally determined structures is rising exponentially. Based on the expectation that protein structures will continue to be determined at an ever-increasing rate, reliable structure prediction schemes will become increasingly valuable, leading to information on protein function and disease for many different proteins. Given known genetic variability and experimentally determined protein structures, can we accurately predict the effects of single amino acid substitutions? An objective assessment of this question would involve comparing predicted and experimentally determined structures, which thus far has not been rigorously performed. The completed research leveraged existing expertise at LLNL in computational and structural biology, as well as significant computing resources, to address this question.« less

Authors:
; ;
Publication Date:
Research Org.:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
898012
Report Number(s):
UCRL-TR-218296
TRN: US200705%%522
DOE Contract Number:
W-7405-ENG-48
Resource Type:
Technical Report
Country of Publication:
United States
Language:
English
Subject:
59 BASIC BIOLOGICAL SCIENCES; AMINO ACIDS; BIOLOGY; CHROMOSOMES; CRYSTALLOGRAPHY; DISEASES; DNA; FORECASTING; GENETIC VARIABILITY; GENETICS; MUTATIONS; NEOPLASMS; NUCLEOTIDES; PROTEIN STRUCTURE; PROTEINS; RESIDUES; RHEUMATIC DISEASES; SORTING; SPECTROSCOPY

Citation Formats

Beernink, P, Barsky, D, and Pesavento, B. Mutations that Cause Human Disease: A Computational/Experimental Approach. United States: N. p., 2006. Web. doi:10.2172/898012.
Beernink, P, Barsky, D, & Pesavento, B. Mutations that Cause Human Disease: A Computational/Experimental Approach. United States. doi:10.2172/898012.
Beernink, P, Barsky, D, and Pesavento, B. Wed . "Mutations that Cause Human Disease: A Computational/Experimental Approach". United States. doi:10.2172/898012. https://www.osti.gov/servlets/purl/898012.
@article{osti_898012,
title = {Mutations that Cause Human Disease: A Computational/Experimental Approach},
author = {Beernink, P and Barsky, D and Pesavento, B},
abstractNote = {International genome sequencing projects have produced billions of nucleotides (letters) of DNA sequence data, including the complete genome sequences of 74 organisms. These genome sequences have created many new scientific opportunities, including the ability to identify sequence variations among individuals within a species. These genetic differences, which are known as single nucleotide polymorphisms (SNPs), are particularly important in understanding the genetic basis for disease susceptibility. Since the report of the complete human genome sequence, over two million human SNPs have been identified, including a large-scale comparison of an entire chromosome from twenty individuals. Of the protein coding SNPs (cSNPs), approximately half leads to a single amino acid change in the encoded protein (non-synonymous coding SNPs). Most of these changes are functionally silent, while the remainder negatively impact the protein and sometimes cause human disease. To date, over 550 SNPs have been found to cause single locus (monogenic) diseases and many others have been associated with polygenic diseases. SNPs have been linked to specific human diseases, including late-onset Parkinson disease, autism, rheumatoid arthritis and cancer. The ability to predict accurately the effects of these SNPs on protein function would represent a major advance toward understanding these diseases. To date several attempts have been made toward predicting the effects of such mutations. The most successful of these is a computational approach called ''Sorting Intolerant From Tolerant'' (SIFT). This method uses sequence conservation among many similar proteins to predict which residues in a protein are functionally important. However, this method suffers from several limitations. First, a query sequence must have a sufficient number of relatives to infer sequence conservation. Second, this method does not make use of or provide any information on protein structure, which can be used to understand how an amino acid change affects the protein. The experimental methods that provide the most detailed structural information on proteins are X-ray crystallography and NMR spectroscopy. However, these methods are labor intensive and currently cannot be carried out on a genomic scale. Nonetheless, Structural Genomics projects are being pursued by more than a dozen groups and consortia worldwide and as a result the number of experimentally determined structures is rising exponentially. Based on the expectation that protein structures will continue to be determined at an ever-increasing rate, reliable structure prediction schemes will become increasingly valuable, leading to information on protein function and disease for many different proteins. Given known genetic variability and experimentally determined protein structures, can we accurately predict the effects of single amino acid substitutions? An objective assessment of this question would involve comparing predicted and experimentally determined structures, which thus far has not been rigorously performed. The completed research leveraged existing expertise at LLNL in computational and structural biology, as well as significant computing resources, to address this question.},
doi = {10.2172/898012},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Wed Jan 11 00:00:00 EST 2006},
month = {Wed Jan 11 00:00:00 EST 2006}
}

Technical Report:

Save / Share:
  • The authors have developed and enhanced a set of tools for fold recognition with hidden Markov models (HMMs), and used these tools effectively in the CASP2 protein structure prediction contest [KKB+97]. HMMs have limitations, and one limitation is that they do not model the long-range pairwise interactions that define the shape of a protein. As such, the authors are working on modeling pairwise interactions to incorporate them into the HMM-based framework. Classical fold recognition methods are based on the premise of distinct pairwise preferences between two given amino acids. The authors have studied these preferences extensively and found that inmore » the general case, this information is limited. Yet by modeling pairwise interactions in context of phylogenetic relationships and by modeling one specific type of contact, the contact between interacting beta strand residues, they have recovered significant information for prediction and analysis of protein structure.« less
  • This is the final report of a three-year, Laboratory-Directed Research and Development (LDRD) project at the Los Alamos National Laboratory (LANL). This program developed project combined Los Alamos' and collaborators' strengths in noninvasive brain imaging and high performance computing to develop potential contributions to the multi-agency Human Brain Project led by the National Institute of Mental Health. The experimental component of the project emphasized the optimization of spatial and temporal resolution of functional brain imaging by combining: (a) structural MRI measurements of brain anatomy; (b) functional MRI measurements of blood flow and oxygenation; and (c) MEG measurements of time-resolved neuronalmore » population currents. The computational component of the project emphasized development of a high-resolution 3-D volumetric model of the brain based on anatomical MRI, in which structural and functional information from multiple imaging modalities can be integrated into a single computational framework for modeling, visualization, and database representation.« less
  • In this project, a computational modeling approach for analyzing flow and ash transport and deposition in filter vessels was developed. An Eulerian-Lagrangian formulation for studying hot-gas filtration process was established. The approach uses an Eulerian analysis of gas flows in the filter vessel, and makes use of the Lagrangian trajectory analysis for the particle transport and deposition. Particular attention was given to the Siemens-Westinghouse filter vessel at Power System Development Facility in Wilsonville in Alabama. Details of hot-gas flow in this tangential flow filter vessel are evaluated. The simulation results show that the rapidly rotation flow in the spacing betweenmore » the shroud and the vessel refractory acts as cyclone that leads to the removal of a large fraction of the larger particles from the gas stream. Several alternate designs for the filter vessel are considered. These include a vessel with a short shroud, a filter vessel with no shroud and a vessel with a deflector plate. The hot-gas flow and particle transport and deposition in various vessels are evaluated. The deposition patterns in various vessels are compared. It is shown that certain filter vessel designs allow for the large particles to remain suspended in the gas stream and to deposit on the filters. The presence of the larger particles in the filter cake leads to lower mechanical strength thus allowing for the back-pulse process to more easily remove the filter cake. A laboratory-scale filter vessel for testing the cold flow condition was designed and fabricated. A laser-based flow visualization technique is used and the gas flow condition in the laboratory-scale vessel was experimental studied. A computer model for the experimental vessel was also developed and the gas flow and particle transport patterns are evaluated.« less
  • Efforts during this first year focused on four areas: (1) searching/summarizing published FTS mechanistic and kinetic studies of FTS reactions on iron catalysts; (2) construction of mass spectrometer-TPD and Berty CSTR reactor systems; (3) preparation and characterization of unsupported iron and alumina-supported iron catalysts at various iron loadings (4) Determination of thermochemical parameters such as binding energies of reactive intermediates, heat of FTS elementary reaction steps, and kinetic parameters such as activation energies, and frequency factors of FTS elementary reaction steps on a number of model surfaces. Literature describing mechanistic and kinetic studies of Fischer-Tropsch synthesis on iron catalysts wasmore » compiled in a draft review. Construction of the mass spectrometer-TPD system is 90% complete and of a Berty CSTR reactor system 98% complete. Three unsupported iron catalysts and three alumina-supported iron catalysts were prepared by nonaqueous-evaporative deposition (NED) or aqueous impregnation (AI) and characterized by chemisorption, BET, extent-of-reduction, XRD, and TEM methods. These catalysts, covering a wide range of dispersions and metal loadings, are well-reduced and relatively thermally stable up to 500-600 C in H{sub 2}, thus ideal for kinetic and mechanistic studies. The alumina-supported iron catalysts will be used for kinetic and mechanistic studies. In the coming year, adsorption/desorption properties, rates of elementary steps, and global reaction rates will be measured for these catalysts, with and without promoters, providing a database for understanding effects of dispersion, metal loading, and support on elementary kinetic parameters and for validation of computational models that incorporate effects of surface structure and promoters. Furthermore, using state-of-the-art self-consistent Density Functional Theory (DFT) methods, we have extensively studied the thermochemistry and kinetics of various elementary steps on three different model surfaces: (1) Fe(110), (2) Fe(110) modified by subsurface C, and (3) Fe surface modified with Pt adatoms. These studies have yielded valuable insights into the reactivity of Fe surfaces for FTS, and provided accurate estimates for the effect of Fe modifiers such as subsurface C and surface Pt.« less