Mehmet Serkan Apaydin

An optimal procedure for gap closing in whole genome shotgun sequencing

Tettelin et. al. proposed a new method for closing the gaps in whole genome shotgun sequencing projects. The method uses a multiplex PCR strategy in order to minimize the time and effort required to sequence the DNA in the missing gaps. This procedure has been used in a number of microbial sequencing projects including Streptococcus pneumoniae and other bacteria. In this paper we describe a theoretical framework for this problem and propose an improved method that guarantees to minimize the number of steps involved in the gap closure procedures. In given particular collection of n/2 DNA fragments we describe a strategy that requires. 0.75 log n work in eight parallel rounds of experiment closely matching a corresponding lower bound 0.5 log of n

Capturing molecular energy landscapes with probabilistic conformational roadmaps

Probabilistic roadmaps are an effective tool to compute the connectivity of the collision-free subset of high-dimensional robot configuration spaces. This paper extends them to capture the pertinent features of continuous functions over high-dimensional spaces. We focus here on computing energetically favorable motions of bio-molecules. A molecule is modeled as an articulated structure moving in an energy field. The set of all its 3D placements is the molecules conformational space, over which the energy field is defined. A probabilistic conformational roadmap (PCR) tries to capture the connectivity of the low-energy subset of a conformational space, in the form of a network of weighted local pathways. The weight of a pathway measures the energetic difficulty for the molecule to move along it. The power of a PCR derives from its ability to compactly encode a large number of energetically favorable molecular pathways, each defined as a sequence of contiguous local pathways. This paper describes general techniques to compute and query PCRs, and presents implementations to study ligand-protein binding and protein folding.

Stochastic Conformational Roadmaps for Computing Ensemble Properties of Molecular Motion

A key intuition behind probabilistic roadmap planners for motion planning is that many collision-free paths potentially exist between two given robot configurations. Hence the connectivity of a robot’s free space can be captured effectively by a network of randomly sampled configurations. In this paper, a similar intuition is exploited to preprocess molecular motion pathways and efficiently compute their ensemble properties, i.e., properties characterizing the average behavior of many pathways. We construct a directed graph, called stochastic conformational roadmap, whose nodes are randomly sampled molecule conformations. A roadmap compactly encodes many molecular motion pathways. Ensemble properties are computed by viewing the roadmap as a Markov chain. A salient feature of this new approach is that it examines all the paths in the roadmap simultaneously, rather than one at a time as classic methods such as Monte Carlo (MC) simulation would do. It also avoids the local-minima problem encountered by the classic methods. Tests of the approach on two important biological problems show that it produces more accurate results and achieves several orders of magnitude reduction in computation time, compared with MC simulation.

A Tabu Search Approach for the NMR Protein Structure-Based Assignment Problem

Nuclear Magnetic Resonance (NMR) (Abbreviations used: NMR, Nuclear Magnetic Resonance; NOE, Nuclear Overhauser Effect; RDC, Residual Dipolar Coupling; PDB, Protein Data Bank; SBA, Structure-Based Assignments; NVR, Nuclear Vector Replacement; BIP, Binary Integer Programming; TS, Tabu Search; QAP, Quadratic Assignment Problem; ff2, the FF Domain 2 of human transcription elongation factor CA150 (RNA polymerase II C-terminal domain interacting protein); SPG, Streptococcal Protein G; hSRI, Human Set2-Rpb1 Interacting Domain; MBP, Maltose Binding Protein; EIN, Amino Terminal Domain of Enzyme I from Escherichia Coli; EM, expectation maximization) Spectroscopy is an experimental technique which exploits the magnetic properties of specific nuclei and enables the study of proteins in solution. The key bottleneck of NMR studies is to map the NMR peaks to corresponding nuclei, also known as the assignment problem. Structure-Based Assignment (SBA) is an approach to solve this computationally challenging problem by using prior information about the protein obtained from a homologous structure. NVR-BIP used the Nuclear Vector Replacement (NVR) framework to model SBA as a binary integer programming problem. In this paper, we prove that this problem is NP-hard and propose a tabu search (TS) algorithm (NVR-TS) equipped with a guided perturbation mechanism to efficiently solve it. NVR-TS uses a quadratic penalty relaxation of NVR-BIP where the violations in the Nuclear Overhauser Effect constraints are penalized in the objective function. Experimental results indicate that our algorithm finds the optimal solution on NVRBIPs data set which consists of seven proteins with 25 templates (31 to 126 residues). Furthermore, it achieves relatively high assignment accuracies on two additional large proteins, MBP and EIN (348 and 243 residues, respectively), which NVR-BIP failed to solve. The executable and the input files are available for download at http://people.sabanciuniv.edu/catay/NVR-TS/NVR-TS.html.

Predicting experimental quantities in protein folding kinetics using stochastic roadmap simulation

This paper presents a new method for studying protein folding kinetics. It uses the recently introduced Stochastic Roadmap Simulation (SRS) method to estimate the transition state ensemble (TSE) and predict the rates and Φ-values for protein folding. The new method was tested on 16 proteins. Comparison with experimental data shows that it estimates the TSE much more accurately than an existing method based on dynamic programming. This leads to better folding-rate predictions. The results on Φ-value predictions are mixed, possibly due to the simple energy model used in the tests. This is the first time that results obtained from SRS have been compared against a substantial amount of experimental data. The success further validates the SRS method and indicates its potential as a general tool for studying protein folding kinetics.

An ant colony optimization approach for solving the nuclear magnetic resonance structure based assignment problem

Nuclear Magnetic Resonance (NMR) Spectroscopy is an important technique that allows determining protein structure in solution. An important problem in protein structure determination using NMR spectroscopy is the mapping of peaks to corresponding amino acids. Structure Based Assignment (SBA) is an approach to solve this problem using a template structure that is homologous to the target. Our previously developed approach NVR-BIP computed the optimal solution for small proteins, but was unable to solve the assignments of large proteins. NVR-TS extended the applicability of the NVR approach for such proteins, however the accuracies varied significantly from run to run. In this paper, we propose NVR-ACO, an Ant Colony Optimization (ACO) based approach to this problem. NVRACO is similar to other ACO algorithms in a way that it also consists of three phases: the construction phase, an optional local search phase and a pheromone update phase. But it has some important differences from other ACO algorithms in terms of solution construction and pheromone update functions and convergence rules. We studied the data set used in NVR-BIP and NVR-TS. Our new method finds optimal solutions for small proteins and achieves perfect assignment on EIN and higher accuracy on MBP compared to NVR-TS. It is also more robust.

Using amino acid typing to improve the accuracy of NMR structure based assignments

Nuclear Magnetic Resonance (NMR1) spectroscopy is an important experimental technique that allows one to study protein structure in solution. An important challenge in NMR protein structure determination is the assignment of NMR peaks to corresponding nuclei. In structure-based assignment (SBA), the aim is to perform the assignments with the help of a homologous protein. NVR-BIP is a tool that uses Nuclear Vector Replacements (NVR) scoring function and binary integer programming to solve SBA problem. In this work, we introduce a method to improve NVR-BIPs assignment accuracy with amino acid typing. We use CRAACK that takes the chemical shifts of C, N and H atoms and returns the possible amino acids along with their confidence scores. We obtain improved assignment accuracies and our results show the effectiveness of integrating amino acid typing with NVR.

Automating unambiguous NOE data usage in NVR for NMR protein structure-based assignments

Nuclear Magnetic Resonance (NMR) Spectroscopy is an important technique that allows determining protein structure in solution. An important problem in protein structure determination using NMR spectroscopy is the mapping of peaks to corresponding amino acids, also known as the assignment problem. Structure-Based Assignment (SBA) is an approach to solve this problem using a template structure that is homologous to the target. Our previously developed approach Nuclear Vector Replacement-Binary Integer Programming (NVR-BIP) computed the optimal solution for small proteins, but was unable to solve the assignments of large proteins. NVR-Ant Colony Optimization (ACO) extended the applicability of the NVR approach for such proteins. One of the input data utilized in these approaches is the Nuclear Overhauser Effect (NOE) data. NOE is an interaction observed between two protons if the protons are located close in space. These protons could be amide protons, protons attached to the alpha-carbon atom in the backbone of the protein, or side chain protons. NVR only uses backbone protons. In this paper, we reformulate the NVR-BIP model to distinguish the type of proton in NOE data and use the corresponding proton coordinates in the extended formulation. In addition, the threshold value over interproton distances is set in a standard manner for all proteins by extracting the NOE upper bound distance information from the data. We also convert NOE intensities into distance thresholds. Our new approach thus handles the NOE data correctly and without manually determined parameters. We accordingly adapt NVR-ACO solution methodology to these changes. Computational results show that our approaches obtain optimal solutions for small proteins. For the large proteins our ant colony optimization-based approach obtains promising results.

Traffic Speed Prediction with Neural Networks

With the increasing interest in creating Smart Cities, traffic speed prediction has attracted more attention in contemporary transportation research. Neural networks have been utilized in many studies to address this problem; yet, they have mainly focused on the short-term prediction while longer forecast horizons are needed for more reliable mobility and route planning. In this work we tackle the medium-term prediction as well as the short-term. We employ feedforward neural networks that combine time series forecasting techniques where the predicted speed values are fed into the network. We train our networks and select the hyper-parameters to minimize the mean absolute error. To test the performance of our method, we consider two multi-segment routes in Istanbul. The speed data are collected from floating cars for every minute over a 5-month horizon. Our computational results showed that accurate predictions can be achieved in medium-term horizon.

Progress in Nuclear Vector Replacement for NMR Protein Structure-Based Assignments

Nuclear Magnetic Resonance (NMR) Spectroscopy is an important technique to obtain structural information of a protein. In this technique, an essential step is the backbone resonance assignment and Structure Based Assignment (SBA) aims to solve this problem with the help of a template structure. Nuclear Vector Replacement (NVR) is an NMR protein SBA program, that takes as input 15 N and H N chemical shifts and unambiguous NOEs, as well as RDCs, HD-exchange and TOCSY data. NVR does not utilize 13 C chemical shifts although this data is widely available for many proteins. In addition, NVR is a proof-of-principle approach and has been run with specific and manually set parameters for some proteins. NA-NVR-ACO [M. Akhmedov, B.Catay and M.S. Apaydin, J. Bioinform. Comput. Biol. 13 (2015) 1550020.] remedies this problem for the NOE data and standardizes NOE usage, while using an ant colony optimization based algorithm. In this paper, we standardize NA-NVR-ACO’s scoring function by using the same parameters for all the proteins and incorporating 13 C α chemical shifts. We also use a larger protein database and state-of-the-art chemical shift prediction tools, SHIFTX2 [B. Han, Y. Liu, S.W. Ginzinger and D.S. Wishart, J. Biomol. NMR 50 (2011) 43–57.] and SPARTA+ [Y. Shen and A. Bax, J. Biomol. NMR 48 (2010) 13–22], to extract the chemical shift statistics. Other practical improvements include automatizing data file preparation and obtaining a degree of reliability for individual peak-amino acid assignments. Our results show that our improvements bring NA-NVR-ACO closer to a practical tool, able to handle a variety of different data types.