Towards Deep Learning Approaches for Quantitative Analysis of High-Throughput DLD

2020 ◽  
Author(s):  
Eric A. Gioe ◽  
Xiaolin Chen ◽  
Jong-Hoon Kim
Keyword(s):  
Deep Learning ◽  
High Throughput ◽  
Lateral Displacement ◽  
Great Promise ◽  
Learning Approaches ◽  
Timely Manner ◽  
Particle Detection ◽  
New Techniques ◽  
Deterministic Lateral Displacement ◽  
Total Particle

Abstract Microfluidics has shown great promise for the sorting or separation of biological cells such as circulating tumor cells since the first studies came out a few decades ago. With recent advances in high-throughput microfluidics, analysis of massive amounts of data needs to be completed in an iterative, timely manner. However, the majority of analysis is either performed manually or through the use of superimposing multiple images to define the flow of the particles, taking a significant amount of time to complete. The objective of the work is to increase the efficiency and repeatability of particle detection in a high-throughput deterministic lateral displacement (DLD) device. The program proposed is in the early stages of development, but shows promise. The average time it takes to analyze over a gigabyte of video is 24.21 seconds. The average percent error of the total particle detection was 21.42%. The assumptions made for the initial version of the program affect the accuracy of the particle in wall detection, so new techniques that do not follow the assumptions will need to be investigated. More work will be completed to implement machine learning or deep learning to assist in the development of DLD devices.

scholarly journals PARROT is a flexible recurrent neural network framework for analysis of large protein datasets

eLife ◽  
2021 ◽  
Vol 10 ◽  
Author(s):  
Daniel Griffith ◽  
Alex S Holehouse
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Deep Learning ◽  
High Throughput ◽  
Recurrent Neural Network ◽  
Transcriptional Activation ◽  
Network Architecture ◽  
Learning Approaches ◽  
Large Protein ◽  
Protein Datasets

The rise of high-throughput experiments has transformed how scientists approach biological questions. The ubiquity of large-scale assays that can test thousands of samples in a day has necessitated the development of new computational approaches to interpret this data. Among these tools, machine learning approaches are increasingly being utilized due to their ability to infer complex nonlinear patterns from high-dimensional data. Despite their effectiveness, machine learning (and in particular deep learning) approaches are not always accessible or easy to implement for those with limited computational expertise. Here we present PARROT, a general framework for training and applying deep learning-based predictors on large protein datasets. Using an internal recurrent neural network architecture, PARROT is capable of tackling both classification and regression tasks while only requiring raw protein sequences as input. We showcase the potential uses of PARROT on three diverse machine learning tasks: predicting phosphorylation sites, predicting transcriptional activation function of peptides generated by high-throughput reporter assays, and predicting the fibrillization propensity of amyloid beta with data generated by deep mutational scanning. Through these examples, we demonstrate that PARROT is easy to use, performs comparably to state-of-the-art computational tools, and is applicable for a wide array of biological problems.

Study of Angle-of-Attack (AoA) for Airfoil in Deterministic Lateral Displacement (DLD)

2019 ◽  
Author(s):  
Kawkab Ahasan ◽  
Jong-Hoon Kim
Keyword(s):  
High Throughput ◽  
Lateral Displacement ◽  
Angle Of Attack ◽  
Critical Diameter ◽  
Control Flow ◽  
Flow Rates ◽  
Deterministic Lateral Displacement ◽  
Potential Applications ◽  
High Reynolds ◽  
Flow Symmetry

Abstract Deterministic lateral displacement (DLD) is a method of inertial size-based particle separation with potential applications in high throughput sample processing, such as the fractionation of blood or the purification of target species like viral particles or circulating tumor cells. Recently, it has been shown that symmetric airfoils with neutral angle-of-attack (AoA) can be used for high-throughput design of DLD device, due to their mitigation of vortex effects and preservation of flow symmetry under high Reynolds number (Re) conditions. While high-Re operation with symmetric airfoils has been established, the effect of AoA for airfoil on the DLD performance has not been characterized. In this study, we present a high-Re investigation with symmetric airfoil-shaped pillars having positive and negative 15 degree AoA. Both positive and negative AoA configurations yield significant flow anisotropy at higher flow rates. The stronger shift of the critical diameter (Dc) was observed with negative AoA, but not in positive AoA device. The most likely contributor may be the growing anisotropy that develops in the AoA device at higher flow rates. This study shows that high-Re DLD design with airfoil shaped pillars requires significant consideration for pillar orientation to control flow symmetry.

scholarly journals Deep Learning Approaches for Protein Structure Prediction

2018 ◽  
Vol 7 (4.5) ◽  
pp. 168
Author(s):  
Khatri Chandni ◽  
Prof. Mrudang Pandya ◽  
Dr. Sunil Jardosh
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
Protein Structure ◽  
Protein Structure Prediction ◽  
Structure Prediction ◽  
Machine Learning Techniques ◽  
Great Promise ◽  
Learning Approaches ◽  
Learning Networks ◽  
Learning Techniques

In recent years, Machine Learning techniques that are based on Deep Learning networks that show a great promise in research          communities.Successful methods for deep learning involve Artificial Neural Networks and Machine Learning. Deep Learning solves severa  problems in bioinformatics. Protein Structure Prediction is one of the most important fields that can be solving using Deep Learning  approaches.These protein are categorized on basis of occurrence of amino acid patterns occur to extract the feature. In these paper aimed to review work based on protein structure prediction solve using Deep Learning Networks. Objective is to review motivate and facilitatethese deep learn the network for predicting protein sequences using Deep Learning. 

scholarly journals Outdoor Plant Segmentation With Deep Learning for High-Throughput Field Phenotyping on a Diverse Wheat Dataset

2022 ◽  
Vol 12 ◽  
Author(s):  
Radek Zenkl ◽  
Radu Timofte ◽  
Norbert Kirchgessner ◽  
Lukas Roth ◽  
Andreas Hund ◽  
...  
Keyword(s):  
Machine Learning ◽  
Deep Learning ◽  
High Throughput ◽  
Developmental Stages ◽  
Agricultural Practices ◽  
Support Vector ◽  
Learning Approaches ◽  
Support Vector Classifier ◽  
Wide Range ◽  
Field Phenotyping

Robust and automated segmentation of leaves and other backgrounds is a core prerequisite of most approaches in high-throughput field phenotyping. So far, the possibilities of deep learning approaches for this purpose have not been explored adequately, partly due to a lack of publicly available, appropriate datasets. This study presents a workflow based on DeepLab v3+ and on a diverse annotated dataset of 190 RGB (350 x 350 pixels) images. Images of winter wheat plants of 76 different genotypes and developmental stages have been acquired throughout multiple years at high resolution in outdoor conditions using nadir view, encompassing a wide range of imaging conditions. Inconsistencies of human annotators in complex images have been quantified, and metadata information of camera settings has been included. The proposed approach achieves an intersection over union (IoU) of 0.77 and 0.90 for plants and soil, respectively. This outperforms the benchmarked machine learning methods which use Support Vector Classifier and/or Random Forrest. The results show that a small but carefully chosen and annotated set of images can provide a good basis for a powerful segmentation pipeline. Compared to earlier methods based on machine learning, the proposed method achieves better performance on the selected dataset in spite of using a deep learning approach with limited data. Increasing the amount of publicly available data with high human agreement on annotations and further development of deep neural network architectures will provide high potential for robust field-based plant segmentation in the near future. This, in turn, will be a cornerstone of data-driven improvement in crop breeding and agricultural practices of global benefit.

Enhancing the Cell Viability in High Throughput Deterministic Lateral Displacement Separation of Circulating Tumor Cells

2019 ◽  
Author(s):  
Arian Aghilinejad ◽  
Christopher Landry ◽  
George Cha ◽  
Xiaolin Chen
Keyword(s):  
Reynolds Number ◽  
Tumor Cells ◽  
Circulating Tumor Cells ◽  
High Throughput ◽  
Cancer Metastasis ◽  
Lateral Displacement ◽  
Shear Stresses ◽  
Label Free ◽  
Deterministic Lateral Displacement ◽  
Reynolds Number Flow

Abstract Cancer is among a major health concerns all over the world. Cancer metastasis, which defines as the migration of malignant cells from original sites to distant organs, is the main reason of death due to cancer and there is growing evidence that Circulating Tumor Cells (CTCs) are responsible for initiating the metastasis. Due to the importance of these bioparticles in biotechnology and medicine, there is a growing interest to study and separate them through different techniques especially microfluidic label-free technologies. One such technology, termed Deterministic Lateral Displacement (DLD) has recently shown promising abilities to separate cells and particles of different sizes. However, DLD is a separation method that takes advantages of the predictable flow laminae of low Reynolds number (Re) fluid flow. In order to achieve higher device throughput, effects of higher Reynolds number flow on DLD device should be studied. Additionally, the higher flow rates would apply higher forces and shear stresses on the cells which threaten the cell’s viability. In this study, employing numerical simulation, the effect of high Re number on DLD device for separating cancer cells has been investigated. Specifically, we conducted force analysis and by focusing on the downstream gap distance between the posts, we improved the device which results in less cell deformation. Our developed numerical model and presented results lay the groundwork for design and fabrication of high-throughput DLD microchips for enhanced separation of CTCs.

Airfoil Deterministic Lateral Displacement for High-Throughput Particle Separation in Viscous Fluids

2020 ◽  
Author(s):  
Brian Senf ◽  
Kawkab Ahasan ◽  
Jong-Hoon Kim
Keyword(s):  
High Throughput ◽  
Biological Sample ◽  
Particle Trajectory ◽  
Lateral Displacement ◽  
Particle Separation ◽  
Reynolds Numbers ◽  
Deionized Water ◽  
Fluid Viscosity ◽  
Deterministic Lateral Displacement ◽  
Low Reynolds Numbers

Abstract Deterministic Lateral Displacement (DLD) is an inertial size-based particle separation technique with great possibilities for use with biological sample preparation. Recently it has been shown that particle shift of a DLD is highly dependent on the Reynolds number. Additionally, particle trajectory has been characterized in a high throughput airfoil array DLD with varying Angle of Attack (AoA) in Deionized water. The AoA can be shifted negatively assisting in particle trajectory increases at low Reynolds numbers. With variations in fluid viscosity, particle trajectories compared to Reynolds value should theoretically have a constant and similar slope. In this work, various viscosities are tested in a DLD with a neutral and negative AoA to eventually characterize non-Newtonian fluids within a DLD. Due to higher viscosities increasing the internal pressure of the device, the negative AoA DLD shows promising results at higher range viscosities due to its ability to shift particles at lower Reynolds numbers.

Sign in / Sign up

Export Citation Format

Share Document