Comprehensive variation discovery in single human genomes.

Complete knowledge of the genetic variation in individual human genomes is a crucial foundation for understanding the etiology of disease. Genetic variation is typically characterized by sequencing individual genomes and comparing reads to a reference. Existing methods do an excellent job of detecting variants in approximately 90% of the human genome; however, calling variants in the remaining 10% of the genome (largely low-complexity sequence and segmental duplications) is challenging. To improve variant calling, we developed a new algorithm, DISCOVAR, and examined its performance on improved, low-cost sequence data. Using a newly created reference set of variants from the finished sequence of 103 randomly chosen fosmids, we find that some standard variant call sets miss up to 25% of variants. We show that the combination of new methods and improved data increases sensitivity by several fold, with the greatest impact in challenging regions of the human genome.

Finished bacterial genomes from shotgun sequence data.

Exceptionally accurate genome reference sequences have proven to be of great value to microbial researchers. Thus, to date, about 1800 bacterial genome assemblies have been "finished" at great expense with the aid of manual laboratory and computational processes that typically iterate over a period of months or even years. By applying a new laboratory design and new assembly algorithm to 16 samples, we demonstrate that assemblies exceeding finished quality can be obtained from whole-genome shotgun data and automated computation. Cost and time requirements are thus dramatically reduced.

Discrete molecular dynamics distinguishes nativelike binding poses from decoys in difficult targets.

Virtual screening is one of the major tools used in computer-aided drug discovery. In structure-based virtual screening, the scoring function is critical to identifying the correct docking pose and accurately predicting the binding affinities of compounds. However, the performance of existing scoring functions has been shown to be uneven for different targets, and some important drug targets have proven especially challenging. In these targets, scoring functions cannot accurately identify the native or near-native binding pose of the ligand from among decoy poses, which affects both the accuracy of the binding affinity prediction and the ability of virtual screening to identify true binders in chemical libraries. Here, we present an approach to discriminating native poses from decoys in difficult targets for which several scoring functions failed to correctly identify the native pose. Our approach employs Discrete Molecular Dynamics simulations to incorporate protein-ligand dynamics and the entropic effects of binding. We analyze a collection of poses generated by docking and find that the residence time of the ligand in the native and nativelike binding poses is distinctly longer than that in decoy poses. This finding suggests that molecular simulations offer a unique approach to distinguishing the native (or nativelike) binding pose from decoy poses that cannot be distinguished using scoring functions that evaluate static structures. The success of our method emphasizes the importance of protein-ligand dynamics in the accurate determination of the binding pose, an aspect that is not addressed in typical docking and scoring protocols.

Cheminformatics meets molecular mechanics: a combined application of knowledge-based pose scoring and physical force field-based hit scoring functions improves the accuracy of structure-based virtual screening.

Poor performance of scoring functions is a well-known bottleneck in structure-based virtual screening (VS), which is most frequently manifested in the scoring functions' inability to discriminate between true ligands vs known nonbinders (therefore designated as binding decoys). This deficiency leads to a large number of false positive hits resulting from VS. We have hypothesized that filtering out or penalizing docking poses recognized as non-native (i.e., pose decoys) should improve the performance of VS in terms of improved identification of true binders. Using several concepts from the field of cheminformatics, we have developed a novel approach to identifying pose decoys from an ensemble of poses generated by computational docking procedures. We demonstrate that the use of target-specific pose (scoring) filter in combination with a physical force field-based scoring function (MedusaScore) leads to significant improvement of hit rates in VS studies for 12 of the 13 benchmark sets from the clustered version of the Database of Useful Decoys (DUD). This new hybrid scoring function outperforms several conventional structure-based scoring functions, including XSCORE::HMSCORE, ChemScore, PLP, and Chemgauss3, in 6 out of 13 data sets at early stage of VS (up 1% decoys of the screening database). We compare our hybrid method with several novel VS methods that were recently reported to have good performances on the same DUD data sets. We find that the retrieved ligands using our method are chemically more diverse in comparison with two ligand-based methods (FieldScreen and FLAP::LBX). We also compare our method with FLAP::RBLB, a high-performance VS method that also utilizes both the receptor and the cognate ligand structures. Interestingly, we find that the top ligands retrieved using our method are highly complementary to those retrieved using FLAP::RBLB, hinting effective directions for best VS applications. We suggest that this integrative VS approach combining cheminformatics and molecular mechanics methodologies may be applied to a broad variety of protein targets to improve the outcome of structure-based drug discovery studies.

Structural basis for µ-opioid receptor binding and activation.

Opioids that stimulate the µ-opioid receptor (MOR1) are the most frequently prescribed and effective analgesics. Here we present a structural model of MOR1. Molecular dynamics simulations show a ligand-dependent increase in the conformational flexibility of the third intracellular loop that couples with the G protein complex. These simulations likewise identified residues that form frequent contacts with ligands. We validated the binding residues using site-directed mutagenesis coupled with radioligand binding and functional assays. The model was used to blindly screen a library of ∼1.2 million compounds. From the 34 compounds predicted to be strong binders, the top three candidates were examined using biochemical assays. One compound showed high efficacy and potency. Post hoc testing revealed this compound to be nalmefene, a potent clinically used antagonist, thus further validating the model. In summary, the MOR1 model provides a tool for elucidating the structural mechanism of ligand-initiated cell signaling and for screening novel analgesics.

Metastability of free cobalt and iron clusters: a possible precursor to bulk ferromagnetism.

Homonuclear cobalt and iron clusters Co(N) and Fe(N) measured in a cryogenic molecular beam exist in two states with distinct magnetic moments (µ), polarizabilities, and ionization potentials, indicating distinct valences. The µ is approximately quantized: µ(N)∼2Nµ(B) in the ground states and µ(N)(*)∼Nµ(B) in the excited states for Co; µ(N)∼3Nµ(B) and µ(N)(*)∼Nµ(B) for Fe. At a large size, the average µ of the two states converges to the bulk value with diminishing ionization potential differences. The experiments suggest localized ferromagnetism in the two states and that itinerant ferromagnetism emerges from their superposition.

Combined application of cheminformatics- and physical force field-based scoring functions improves binding affinity prediction for CSAR data sets.

The curated CSAR-NRC benchmark sets provide valuable opportunity for testing or comparing the performance of both existing and novel scoring functions. We apply two different scoring functions, both independently and in combination, to predict the binding affinity of ligands in the CSAR-NRC data sets. One reported here for the first time employs multiple chemical-geometrical descriptors of the protein-ligand interface to develop Quantitative Structure Binding Affinity Relationships (QSBAR) models. These models are then used to predict binding affinity of ligands in the external data set. Second is a physical force field-based scoring function, MedusaScore. We show that both individual scoring functions achieve statistically significant prediction accuracies with the squared correlation coefficient (R(2)) between the actual and predicted binding affinity of 0.44/0.53 (Set1/Set2) with QSBAR models and 0.34/0.47 (Set1/Set2) with MedusaScore. Importantly, we find that the combination of QSBAR models and MedusaScore into consensus scoring function affords higher prediction accuracy than any of the contributing methods achieving R(2) values of 0.45/0.58 (Set1/Set2). Furthermore, we identify several chemical features and noncovalent interactions that may be responsible for the inaccurate prediction of binding affinity for several ligands by the scoring functions employed in this study.

Fingerprint-based structure retrieval using electron density.

We present a computational approach that can quickly search a large protein structural database to identify structures that fit a given electron density, such as determined by cryo-electron microscopy. We use geometric invariants (fingerprints) constructed using 3D Zernike moments to describe the electron density, and reduce the problem of fitting of the structure to the electron density to simple fingerprint comparison. Using this approach, we are able to screen the entire Protein Data Bank and identify structures that fit two experimental electron densities determined by cryo-electron microscopy.

Computational evaluation of protein stability change upon mutations.

When designing a mutagenesis experiment, it is often crucial to estimate the stability change of proteins induced by mutations (Delta DG). Despite the recent advances in computational methods, it is still challenging to estimate D DG quickly and accurately. We recently developed the Eris protocols for in silico evaluation of the Delta DG. Starting from the tertiary structure of the wide-type protein, the Eris protocols can model the structure of the mutant protein and estimate Delta DG using the structure models. The Eris protocols not only efficiently optimize the side chains conformations, taking advantage of a fast rotamer-based searching algorithm, but also allow protein backbone flexibility during the modeling. As a result, the Eris protocols effectively resolve steric clashes induced by certain mutations and have more accurate Delta DG predictions than a fixed-backbone approach. We discuss the general aspects of computational Delta DG estimations and discuss in detail the principles and methodologies of the Eris protocols.

Rapid flexible docking using a stochastic rotamer library of ligands.

Existing flexible docking approaches model the ligand and receptor flexibility either separately or in a loosely coupled manner, which captures the conformational changes inefficiently. Here, we propose a flexible docking approach, MedusaDock, which models both ligand and receptor flexibility simultaneously with sets of discrete rotamers. We developed an algorithm to build the ligand rotamer library "on-the-fly" during docking simulations. MedusaDock benchmarks demonstrate a rapid sampling efficiency and high prediction accuracy in both self- (to the cocrystallized state) and cross-docking (to a state cocrystallized with a different ligand), the latter of which mimics the virtual screening procedure in computational drug discovery. We also perform a virtual screening test of four flexible kinase targets, including cyclin-dependent kinase 2, vascular endothelial growth factor receptor 2, HIV reverse transcriptase, and HIV protease. We find significant improvements of virtual screening enrichments when compared to rigid-receptor methods. The predictive power of MedusaDock in cross-docking and preliminary virtual-screening benchmarks highlights the importance to model both ligand and receptor flexibility simultaneously in computational docking.

Computational design of a PAK1 binding protein.

We describe a computational protocol, called DDMI, for redesigning scaffold proteins to bind to a specified region on a target protein. The DDMI protocol is implemented within the Rosetta molecular modeling program and uses rigid-body docking, sequence design, and gradient-based minimization of backbone and side-chain torsion angles to design low-energy interfaces between the scaffold and target protein. Iterative rounds of sequence design and conformational optimization were needed to produce models that have calculated binding energies that are similar to binding energies calculated for native complexes. We also show that additional conformation sampling with molecular dynamics can be iterated with sequence design to further lower the computed energy of the designed complexes. To experimentally test the DDMI protocol, we redesigned the human hyperplastic discs protein to bind to the kinase domain of p21-activated kinase 1 (PAK1). Six designs were experimentally characterized. Two of the designs aggregated and were not characterized further. Of the remaining four designs, three bound to the PAK1 with affinities tighter than 350 muM. The tightest binding design, named Spider Roll, bound with an affinity of 100 muM. NMR-based structure prediction of Spider Roll based on backbone and (13)C(beta) chemical shifts using the program CS-ROSETTA indicated that the architecture of human hyperplastic discs protein is preserved. Mutagenesis studies confirmed that Spider Roll binds the target patch on PAK1. Additionally, Spider Roll binds to full-length PAK1 in its activated state but does not bind PAK1 when it forms an auto-inhibited conformation that blocks the Spider Roll target site. Subsequent NMR characterization of the binding of Spider Roll to PAK1 revealed a comparably small binding 'on-rate' constant (<<10(5) M(-1) s(-1)). The ability to rationally design the site of novel protein-protein interactions is an important step towards creating new proteins that are useful as therapeutics or molecular probes.

Fast screening of protein surfaces using geometric invariant fingerprints.

We develop a rapid and efficient method for the comparison of protein local surface similarities using geometric invariants (fingerprints). By combining fast fingerprint comparison with explicit alignment, we successfully screen the entire Protein Data Bank for proteins that possess local surface similarities. Our method is independent of sequence and fold similarities, and has potential application to protein structure annotation and protein-protein interface design.

MedusaScore: an accurate force field-based scoring function for virtual drug screening.

Virtual screening is becoming an important tool for drug discovery. However, the application of virtual screening has been limited by the lack of accurate scoring functions. Here, we present a novel scoring function, MedusaScore, for evaluating protein-ligand binding. MedusaScore is based on models of physical interactions that include van der Waals, solvation, and hydrogen bonding energies. To ensure the best transferability of the scoring function, we do not use any protein-ligand experimental data for parameter training. We then test the MedusaScore for docking decoy recognition and binding affinity prediction and find superior performance compared to other widely used scoring functions. Statistical analysis indicates that one source of inaccuracy of MedusaScore may arise from the unaccounted entropic loss upon ligand binding, which suggests avenues of approach for further MedusaScore improvement.

Protein folding: then and now.

Over the past three decades the protein folding field has undergone monumental changes. Originally a purely academic question, how a protein folds has now become vital in understanding diseases and our abilities to rationally manipulate cellular life by engineering protein folding pathways. We review and contrast past and recent developments in the protein folding field. Specifically, we discuss the progress in our understanding of protein folding thermodynamics and kinetics, the properties of evasive intermediates, and unfolded states. We also discuss how some abnormalities in protein folding lead to protein aggregation and human diseases.

Modeling backbone flexibility improves protein stability estimation.

In designing mutagenesis experiments, it is often crucial to know how certain mutations will affect the structure and thermodynamic stability of the protein. Here, we present a methodology, Eris, to efficiently and accurately compute the stability changes of proteins upon mutations using our protein-modeling suite, Medusa. We evaluate the stability changes upon mutations for 595 mutants from five structurally unrelated proteins, and find significant correlations between the predicted and experimental results. For cases when the high-resolution protein structure is not available, we find that better predictions are obtained by backbone structure prerelaxation. The advantage of our approach is that it is based on physical descriptions of atomic interactions, and does not rely on parameter training with available experimental protein stability data. Unlike other methods, Eris also models the backbone flexibility, thereby allowing for determination of the mutation-induced backbone conformational changes. Eris is freely available via the web server at http://eris.dokhlab.org.

Magnetic enhancement in cobalt-manganese alloy clusters.

Magnetic moments of Co(N)Mn(M) and Co(N)V(M) clusters (N < or = 60; M < or = N/3) are measured in molecular beams using the Stern-Gerlach deflection method. Surprisingly, the per atom average moments of Co(N)Mn(M) clusters are found to increase with Mn concentration, in contrast to bulk CoMn. The enhancement with Mn doping is found to be independent of cluster size and composition in the size range studied. Meanwhile, Co(N)V(M) clusters show reduction of average moments with increasing V doping, consistent with what is expected in bulk CoV. The results are discussed within the virtual bound states model.

Magnetic moments and adiabatic magnetization of free cobalt clusters.

Magnetizations and magnetic moments of free cobalt clusters Co(N) (12 < N < 200) in a cryogenic (25 K < or = T < or = 100 K) molecular beam were determined from Stern-Gerlach deflections. All clusters preferentially deflect in the direction of the increasing field and the average magnetization resembles the Langevin function for all cluster sizes even at low temperatures. We demonstrate in the avoided crossing model that the average magnetization may result from adiabatic processes of rotating and vibrating clusters in the magnetic field and that spin relaxation is not involved. This resolves a long-standing problem in the interpretation of cluster beam deflection experiments with implications for nanomagnetic systems in general.

Spin uncoupling in free Nb clusters: support for nascent superconductivity.

Molecular beam Stern-Gerlach deflection measurements on Nb clusters (Nb(N), N<100) show that at very low temperatures the odd-N clusters deflect due to a single unpaired spin that is uncoupled from the cluster. At higher temperatures the spin is coupled and no deflections are observed. Spin uncoupling occurs concurrently with the transition to the recently found ferroelectric state, which has superconductor characteristics [Science 300, 1265 (2003)]]. Spin uncoupling (also seen in V, Ta, and Al clusters) is analogous to the reduction of spin-relaxation rates observed in bulk superconductors below T(c).

Ferroelectricity in free niobium clusters.

Electric deflections of gas-phase, cryogenically cooled, neutral niobium clusters [NbN; number of atoms (N) = 2 to 150, temperature (T) = 20to 300kelvin], measured in molecular beams, show that cold clusters may attain an anomalous component with very large electric dipole moments. In contrast, room-temperature measurements show normal metallic polarizabilities. Characteristic energies kBTG(N) [Boltzmann constant kB times a transition temperature TG(N)] are identified, below which the ferroelectric-like state develops. Generally, TG decreases [110 > TG(N) > 10K] as N increases, with pronounced even-odd alternations for N > 38. This new state of metallic matter may be related to bulk superconductivity.