scholarly journals Large-scale simulation of biomembranes: bringing realistic kinetics to coarse-grained models

2019 ◽  
Author(s):  
Mohsen Sadeghi ◽  
Frank Noé
Keyword(s):  
Large Scale ◽  
Stochastic Dynamics ◽  
Computational Cost ◽  
Coarse Grained ◽  
Fluctuation Spectrum ◽  
Biologically Relevant ◽  
Membrane Models ◽  
Kinetic Effects ◽  
The Cost ◽  
Low Computational Cost

Biomembranes are two-dimensional assemblies of phospholipids that are only a few nanometres thick, but form micrometer-sized structures vital to cellular function. Explicit modelling of biologically relevant membrane systems is computationally expensive, especially when the large number of solvent particles and slow membrane kinetics are taken into account. While highly coarse-grained solvent-free models are available to study equilibrium behaviour of membranes, their efficiency comes at the cost of sacrificing realistic kinetics, and thereby the ability to predict pathways and mechanisms of membrane processes. Here, we present a framework for integrating coarse-grained membrane models with anisotropic stochastic dynamics and continuum-based hydrodynamics, allowing us to simulate large biomembrane systems with realistic kinetics at low computational cost. This paves the way for whole-cell simulations that still include nanometer/nanosecond spatiotemporal resolutions. As a demonstration, we obtain and verify fluctuation spectrum of a full-sized human red blood cell in a 150-milliseconds-long single trajectory. We show how the kinetic effects of different cytoplasmic viscosities can be studied with such a simulation, with predictions that agree with single-cell experimental observations.

scholarly journals AlphaImpute2: Fast and accurate pedigree and population based imputation for hundreds of thousands of individuals in livestock populations

2020 ◽  
Author(s):  
Andrew Whalen ◽  
John M Hickey
Keyword(s):  
Large Scale ◽  
Computational Cost ◽  
Population Based ◽  
High Density ◽  
Low Density ◽  
Missing Genotypes ◽  
The Cost ◽  
Iterative Peeling ◽  
Burrows Wheeler Transform ◽  
Number Of Individuals

AbstractIn this paper we present a new imputation algorithm, AlphaImpute2, which performs fast and accurate pedigree and population based imputation for livestock populations of hundreds of thousands of individuals. Genetic imputation is a tool used in genetics to decrease the cost of genotyping a population, by genotyping a small number of individuals at high-density and the remaining individuals at low-density. Shared haplotype segments between the high-density and low-density individuals can then be used to fill in the missing genotypes of the low-density individuals. As the size of genetics datasets have grown, the computational cost of performing imputation has increased, particularly in agricultural breeding programs where there might be hundreds of thousands of genotyped individuals. To address this issue, we present a new imputation algorithm, AlphaImpute2, which performs population imputation by using a particle based approximation to the Li and Stephens which exploits the Positional Burrows Wheeler Transform, and performs pedigree imputation using an approximate version of multi-locus iterative peeling. We tested AlphaImpute2 on four simulated datasets designed to mimic the pedigrees found in a real pig breeding program. We compared AlphaImpute2 to AlphaImpute, AlphaPeel, findhap version 4, and Beagle 5.1. We found that AlphaImpute2 had the highest accuracy, with an accuracy of 0.993 for low-density individuals on the pedigree with 107,000 individuals, compared to an accuracy of 0.942 for Beagle 5.1, 0.940 for AlphaImpute, and 0.801 for findhap. AlphaImpute2 was also the fastest software tested, with a runtime of 105 minutes a pedigree of 107,000 individuals and 5,000 markers was 105 minutes, compared to 190 minutes for Beagle 5.1, 395 minutes for findhap, and 7,859 minutes AlphaImpute. We believe that AlphaImpute2 will enable fast and accurate large scale imputation for agricultural populations as they scale to hundreds of thousands or millions of genotyped individuals.

Designing production-friendly machine learning

2021 ◽  
Vol 14 (13) ◽  
pp. 3420-3420
Author(s):  
Matei Zaharia
Keyword(s):  
Machine Learning ◽  
Open Source ◽  
Large Scale ◽  
Question Answering ◽  
Failure Modes ◽  
Computational Cost ◽  
Language Models ◽  
Software Systems ◽  
Resource Cost ◽  
Low Computational Cost

Building production ML applications is difficult because of their resource cost and complex failure modes. I will discuss these challenges from two perspectives: the Stanford DAWN Lab and experience with large-scale commercial ML users at Databricks. I will then present two emerging ideas to help address these challenges. The first is "ML platforms", an emerging class of software systems that standardize the interfaces used in ML applications to make them easier to build and maintain. I will give a few examples, including the open-source MLflow system from Databricks [3]. The second idea is models that are more "production-friendly" by design. As a concrete example, I will discuss retrieval-based NLP models such as Stanford's ColBERT [1, 2] that query documents from an updateable corpus to perform tasks such as question-answering, which gives multiple practical advantages, including low computational cost, high interpretability, and very fast updates to the model's "knowledge". These models are an exciting alternative to large language models such as GPT-3.

scholarly journals Numerical solution of large scale Hartree-Fock-Bogoliubov equations

2021 ◽  
Author(s):  
Lin Lin ◽  
Xiaojie Wu
Keyword(s):  
Large Scale ◽  
Computational Cost ◽  
Hamiltonian Matrix ◽  
Special Focus ◽  
Matrix Size ◽  
Hartree Fock ◽  
Starting Point ◽  
The Matrix ◽  
The Cost ◽  
Bogoliubov Equations

The Hartree-Fock-Bogoliubov (HFB) theory is the starting point for treating superconducting systems. However, the computational cost for solving large scale HFB equations can be much larger than that of the Hartree-Fock equations, particularly when the Hamiltonian matrix is sparse, and the number of electrons $N$ is relatively small compared to the matrix size $N_{b}$. We first provide a concise and relatively self-contained review of the HFB theory for general finite sized quantum systems, with special focus on the treatment of spin symmetries from a linear algebra perspective. We then demonstrate that the pole expansion and selected inversion (PEXSI) method can be particularly well suited for solving large scale HFB equations. For a Hubbard-type Hamiltonian, the cost of PEXSI is at most $\Or(N_b^2)$ for both gapped and gapless systems, which can be significantly faster than the standard cubic scaling diagonalization methods. We show that PEXSI can solve a two-dimensional Hubbard-Hofstadter model with $N_b$ up to $2.88\times 10^6$, and the wall clock time is less than $100$ s using $17280$ CPU cores. This enables the simulation of physical systems under experimentally realizable magnetic fields, which cannot be otherwise simulated with smaller systems.

scholarly journals Topology optimization subject to additive manufacturing constraints

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Moritz Ebeling-Rump ◽  
Dietmar Hömberg ◽  
Robert Lasarzik ◽  
Thomas Petzold
Keyword(s):  
Topology Optimization ◽  
Additive Manufacturing ◽  
Computational Cost ◽  
Ginzburg Landau ◽  
First Order ◽  
Manufacturing Constraint ◽  
External Forces ◽  
The Cost ◽  
Low Computational Cost ◽  
The Ideal

AbstractIn topology optimization the goal is to find the ideal material distribution in a domain subject to external forces. The structure is optimal if it has the highest possible stiffness. A volume constraint ensures filigree structures, which are regulated via a Ginzburg–Landau term. During 3D printing overhangs lead to instabilities. As a remedy an additive manufacturing constraint is added to the cost functional. First order optimality conditions are derived using a formal Lagrangian approach. With an Allen-Cahn interface propagation the optimization problem is solved iteratively. At a low computational cost the additive manufacturing constraint brings about support structures, which can be fine tuned according to demands and increase stability during the printing process.

scholarly journals Video Based Person Re-Identification Through Selective Knowledge Distillation

2019 ◽  
pp. 699-703
Author(s):  
Gudavalli Sai Abhilash ◽  
Kantheti Rajesh ◽  
Jangam Dileep Shaleem ◽  
Grandi Sai Sarath ◽  
Palli R Krishna Prasad
Keyword(s):  
Computational Cost ◽  
Low Complexity ◽  
Fine Tuning ◽  
Low Resolution ◽  
Facial Cues ◽  
Significant Performance ◽  
Knowledge Distillation ◽  
Computational Resources ◽  
The Cost ◽  
Low Computational Cost

The creation and deployment of face recognition models need to identify low-resolution faces with extremely low computational cost. To address this problem, a feasible solution is compressing a complex face model to achieve higher speed and lower memory at the cost of minimal performance drop. Inspired by that, this paper proposes a learning approach to recognize low-resolution faces via selective knowledge distillation in live video. In this approach, a two-stream convolution neural network (CNN) is first initialized to recognize high-resolution faces and resolution-degraded faces with a teacher stream and a student stream, respectively. The teacher stream is represented by a complex CNN for high-accuracy recognition, and the student stream is represented by a much simpler CNN for low-complexity recognition. To avoid significant performance drop at the student stream, we then selectively distil the most informative facial features from the teacher stream by solving a sparse graph optimization problem, which are then used to regularize the fine- tuning process of the student stream. In this way, the student stream is actually trained by simultaneously handling two tasks with limited computational resources approximating the most informative facial cues via feature regression, and recovering the missing facial cues via low-resolution face classification.

scholarly journals Capturing Protein-Ligand Recognition Pathways in Coarse-grained Simulation

2019 ◽  
Author(s):  
Bhupendra R. Dandekar ◽  
Jagannath Mondal
Keyword(s):  
Large Scale ◽  
Computational Cost ◽  
Md Simulations ◽  
Coarse Grained ◽  
Attractive Alternative ◽  
Complex Process ◽  
Potential Binding ◽  
Allosteric Sites ◽  
Potential Binding Site ◽  
Complete Process

AbstractProtein-substrate recognition is highly dynamic and complex process in nature. A key approach in deciphering the mechanism underlying the recognition process is to capture the kinetic process of substrate in its act of binding to its designated protein cavity. Towards this end, microsecond long atomistic molecular dynamics (MD) simulation has recently emerged as a popular method of choice, due its ability to record these events at high spatial and temporal resolution. However, success in this approach comes at an exorbitant computational cost. Here we demonstrate that coarse grained models of protein, when systematically optimised to maintain its tertiary fold, can capture the complete process of spontaneous protein-ligand binding from bulk media to cavity, within orders of magnitude shorter wall clock time compared to that of all-atom MD simulations. The simulated and crystallographic binding pose are in excellent agreement. We find that the exhaustive sampling of ligand exploration in protein and solvent, harnessed by coarse-grained simulation at a frugal computational cost, in combination with Markov state modelling, leads to clearer mechanistic insights and discovery of novel recognition pathways. The result is successfully validated against three popular protein-ligand systems. Overall, the approach provides an affordable and attractive alternative of all-atom simulation and promises a way-forward for replacing traditional docking based small molecule discovery by high-throughput coarse-grained simulation for searching potential binding site and allosteric sites. This also provides practical avenues for first-hand exploration of bio-molecular recognition processes in large-scale biological systems, otherwise inaccessible in all-atom simulations.

scholarly journals Impact of the reduced speed of light approximation on ionization front velocities in cosmological simulations of the epoch of reionization

2019 ◽  
Vol 622 ◽  
pp. A142 ◽  
Author(s):  
Nicolas Deparis ◽  
Dominique Aubert ◽  
Pierre Ocvirk ◽  
Jonathan Chardin ◽  
Joseph Lewis
Keyword(s):  
Large Scale ◽  
Computational Cost ◽  
Ionization Front ◽  
Speed Of Light ◽  
Low Velocity ◽  
Cosmological Simulations ◽  
Ionization Fronts ◽  
Minimal Velocity ◽  
Two Stages ◽  
The Cost

Context. Coupled radiative-hydrodynamics simulations of the epoch of reionization aim to reproduce the propagation of ionization fronts during the transition before the overlap of HII regions. Many of these simulations use moment-based methods to track radiative transfer processes using explicit solvers and are therefore subject to strict stability conditions regarding the speed of light, which implies a great computational cost. The cost can be reduced by assuming a reduced speed of light, and this approximation is now widely used to produce large-scale simulations of reionization. Aims. We measure how ionization fronts propagate in simulations of the epoch of reionization. In particular, we want to distinguish between the different stages of the fronts’ progression into the intergalactic medium. We also investigate how these stages and their properties are impacted by the choice of a reduced speed of light. Methods. We introduce a new method for estimating and comparing the ionization front speeds based on maps of the reionization redshifts. We applied it to a set of cosmological simulations of the reionization using a set of reduced speeds of light, and measured the evolution of the ionization front speeds during the reionization process. We only considered models where the reionization is driven by the sources created within the simulations, without potential contributions of an external homogeneous ionizing background. Results. We find that ionization fronts progress via a two-stage process, the first stage at low velocity as the fronts emerge from high density regions and a second later stage just before the overlap, during which front speeds increase close to the speed of light. For example, using a set of small 8 Mpc h−3 simulations, we find that a minimal velocity of 0.3c is able to model these two stages in this specific context without significant impact. Values as low as 0.05c can model the first low velocity stage, but limit the acceleration at later times. Lower values modify the distribution of front speeds at all times. Using another set of simulations with larger 64 Mpc h−3 volumes that better account for distant sources, we find that reduced speed of light has a greater impact on reionization times and front speeds in underdense regions that are reionized at late times and swept by radiation produced by distant sources. Conversely, the same quantities measured in dense regions with slow fronts are less sensitive to c∼ values. While the discrepancies introduced by reduced speed of light could be reduced by the inclusion of an additional UV background, we expect these conclusions to be robust in the case of simulations with reionizations driven by inner sources.

scholarly journals Low Computational-cost Cell Detection Method for Calcium Imaging Data

2018 ◽  
Author(s):  
Tsubasa Ito ◽  
Keisuke Ota ◽  
Kanako Ueno ◽  
Yasuhiro Oisi ◽  
Chie Matsubara ◽  
...  
Keyword(s):  
Calcium Imaging ◽  
Large Scale ◽  
Computational Cost ◽  
Image Data ◽  
Detection Methods ◽  
Cell Detection ◽  
Imaging Data ◽  
Huge Amount ◽  
Detection Algorithms ◽  
Low Computational Cost

AbstractThe rapid progress of calcium imaging has reached a point where the activity of tens of thousands of cells can be recorded simultaneously. However, the huge amount of data in such records makes it difficult to carry out cell detection manually. Consequently, because the cell detection is the first step of multicellular data analysis, there is a pressing need for automatic cell detection methods for large-scale image data. Automatic cell detection algorithms have been pioneered by a handful of research groups. Such algorithms, however, assume a conventional field of view (FOV) (i.e. 512 × 512 pixels) and need a significantly higher computational power for a wider FOV to work within a practical period of time. To overcome this issue, we propose a method called low computational-cost cell detection (LCCD), which can complete its processing even on the latest ultra-large FOV data within a practical period of time. We compared it with two previously proposed methods, constrained non-negative matrix factorization (CNMF) and Suite2P. We found that LCCD makes it possible to detect cells from a huge-amount of high-density imaging data within a shorter period of time and with an accuracy comparable to or better than those of CNMF and Suite2P.

Sign in / Sign up

Export Citation Format

Share Document