scholarly journals Classifying T cell activity in autofluorescence intensity images with convolutional neural networks

2019 ◽  
Author(s):  
Zijie J. Wang ◽  
Alex J. Walsh ◽  
Melissa C. Skala ◽  
Anthony Gitter
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
T Cells ◽  
T Cell ◽  
Convolutional Neural Networks ◽  
Computational Cost ◽  
Cell Activity ◽  
Fine Tuning ◽  
Cell Classification ◽  
Intensity Images

ABSTRACTThe importance of T cells in immunotherapy has motivated developing technologies to better characterize T cells and improve therapeutic efficacy. One specific objective is assessing antigen-induced T cell activation because only functionally active T cells are capable of killing the desired targets. Autofluorescence imaging can distinguish T cell activity states of individual cells in a non-destructive manner by detecting endogenous changes in metabolic co-enzymes such as NAD(P)H. However, recognizing robust patterns of T cell activity is computationally challenging in the absence of exogenous labels or information-rich autofluorescence lifetime measurements. We demonstrate that advanced machine learning can accurately classify T cell activity from NAD(P)H intensity images and that those image-based signatures transfer across human donors. Using a dataset of 8,260 cropped single-cell images from six donors, we meticulously evaluate multiple machine learning models. These range from traditional models that represent images using summary statistics or extract image features with CellProfiler to deep convolutional neural networks (CNNs) pre-trained on general non-biological images. Adapting pre-trained CNNs for the T cell activity classification task provides substantially better performance than traditional models or a simple CNN trained with the autofluorescence images alone. Visualizing the images with dimension reduction provides intuition into why the CNNs achieve higher accuracy than other approaches. However, we observe that fine-tuning all layers of the pre-trained CNN does not provide a classification performance boost commensurate with the additional computational cost. Our software detailing our image processing and model training pipeline is available as Jupyter notebooks at https://github.com/gitter-lab/t-cell-classification.

Nanosecond Photodynamics Simulations of a Cis-Trans Isomerization Are Enabled by Machine Learning

2020 ◽  
Author(s):  
Jingbai Li ◽  
Patrick Reiser ◽  
André Eberhard ◽  
Pascal Friederich ◽  
Steven Lopez
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Excited State ◽  
Adaptive Sampling ◽  
Computational Cost ◽  
Ground Truth ◽  
Absolute Error ◽  
Photochemical Reactions ◽  
Computational Techniques ◽  
Full Potential

<p>Photochemical reactions are being increasingly used to construct complex molecular architectures with mild and straightforward reaction conditions. Computational techniques are increasingly important to understand the reactivities and chemoselectivities of photochemical isomerization reactions because they offer molecular bonding information along the excited-state(s) of photodynamics. These photodynamics simulations are resource-intensive and are typically limited to 1–10 picoseconds and 1,000 trajectories due to high computational cost. Most organic photochemical reactions have excited-state lifetimes exceeding 1 picosecond, which places them outside possible computational studies. Westermeyr <i>et al.</i> demonstrated that a machine learning approach could significantly lengthen photodynamics simulation times for a model system, methylenimmonium cation (CH<sub>2</sub>NH<sub>2</sub><sup>+</sup>).</p><p>We have developed a Python-based code, Python Rapid Artificial Intelligence <i>Ab Initio</i> Molecular Dynamics (PyRAI<sup>2</sup>MD), to accomplish the unprecedented 10 ns <i>cis-trans</i> photodynamics of <i>trans</i>-hexafluoro-2-butene (CF<sub>3</sub>–CH=CH–CF<sub>3</sub>) in 3.5 days. The same simulation would take approximately 58 years with ground-truth multiconfigurational dynamics. We proposed an innovative scheme combining Wigner sampling, geometrical interpolations, and short-time quantum chemical trajectories to effectively sample the initial data, facilitating the adaptive sampling to generate an informative and data-efficient training set with 6,232 data points. Our neural networks achieved chemical accuracy (mean absolute error of 0.032 eV). Our 4,814 trajectories reproduced the S<sub>1</sub> half-life (60.5 fs), the photochemical product ratio (<i>trans</i>: <i>cis</i> = 2.3: 1), and autonomously discovered a pathway towards a carbene. The neural networks have also shown the capability of generalizing the full potential energy surface with chemically incomplete data (<i>trans</i> → <i>cis</i> but not <i>cis</i> → <i>trans</i> pathways) that may offer future automated photochemical reaction discoveries.</p>

TLBO-FLN: Teaching-Learning Based Optimization of Functional Link Neural Networks for Stock Closing Price Prediction

2020 ◽  
Vol 10 (4) ◽  
pp. 522-532 ◽  
Author(s):  
Sarat Chandra Nayak ◽  
Subhranginee Das ◽  
Mohammad Dilsad Ansari
Keyword(s):  
Neural Networks ◽  
Computational Cost ◽  
Optimization Techniques ◽  
Fine Tuning ◽  
Functional Link ◽  
Price Prediction ◽  
Closing Price ◽  
Teaching Learning Based Optimization ◽  
Artificial Neural ◽  
Teaching Learning

Background and Objective: Stock closing price prediction is enormously complicated. Artificial Neural Networks (ANN) are excellent approximation algorithms applied to this area. Several nature-inspired evolutionary optimization techniques are proposed and used in the literature to search the optimum parameters of ANN based forecasting models. However, most of them need fine-tuning of several control parameters as well as algorithm specific parameters to achieve optimal performance. Improper tuning of such parameters either leads toward additional computational cost or local optima. Methods: Teaching Learning Based Optimization (TLBO) is a newly proposed algorithm which does not necessitate any parameters specific to it. The intrinsic capability of Functional Link Artificial Neural Network (FLANN) to recognize the multifaceted nonlinear relationship present in the historical stock data made it popular and got wide applications in the stock market prediction. This article presents a hybrid model termed as Teaching Learning Based Optimization of Functional Neural Networks (TLBO-FLN) by combining the advantages of both TLBO and FLANN. Results and Conclusion: The model is evaluated by predicting the short, medium, and long-term closing prices of four emerging stock markets. The performance of the TLBO-FLN model is measured through Mean Absolute Percentage of Error (MAPE), Average Relative Variance (ARV), and coefficient of determination (R2); compared with that of few other state-of-the-art models similarly trained and found superior.

scholarly journals Anti-tumor effects of RTX-240: an engineered red blood cell expressing 4-1BB ligand and interleukin-15

2021 ◽  
Author(s):  
Shannon L. McArdel ◽  
Anne-Sophie Dugast ◽  
Maegan E. Hoover ◽  
Arjun Bollampalli ◽  
Enping Hong ◽  
...  
Keyword(s):  
T Cells ◽  
T Cell ◽  
Blood Cell ◽  
Nk Cells ◽  
Red Blood Cell ◽  
Safety Profile ◽  
Nk Cell ◽  
Cell Activity ◽  
In Vivo Studies

AbstractRecombinant agonists that activate co-stimulatory and cytokine receptors have shown limited clinical anticancer utility, potentially due to narrow therapeutic windows, the need for coordinated activation of co-stimulatory and cytokine pathways and the failure of agonistic antibodies to recapitulate signaling by endogenous ligands. RTX-240 is a genetically engineered red blood cell expressing 4-1BBL and IL-15/IL-15Rα fusion (IL-15TP). RTX-240 is designed to potently and simultaneously stimulate the 4-1BB and IL-15 pathways, thereby activating and expanding T cells and NK cells, while potentially offering an improved safety profile through restricted biodistribution. We assessed the ability of RTX-240 to expand and activate T cells and NK cells and evaluated the in vivo efficacy, pharmacodynamics and tolerability using murine models. Treatment of PBMCs with RTX-240 induced T cell and NK cell activation and proliferation. In vivo studies using mRBC-240, a mouse surrogate for RTX-240, revealed biodistribution predominantly to the red pulp of the spleen, leading to CD8 + T cell and NK cell expansion. mRBC-240 was efficacious in a B16-F10 melanoma model and led to increased NK cell infiltration into the lungs. mRBC-240 significantly inhibited CT26 tumor growth, in association with an increase in tumor-infiltrating proliferating and cytotoxic CD8 + T cells. mRBC-240 was tolerated and showed no evidence of hepatic injury at the highest feasible dose, compared with a 4-1BB agonistic antibody. RTX-240 promotes T cell and NK cell activity in preclinical models and shows efficacy and an improved safety profile. Based on these data, RTX-240 is now being evaluated in a clinical trial.

scholarly journals Characterization of human T cells reactive with the Mycoplasma arthritidis-derived superantigen (MAM): generation of a monoclonal antibody against V beta 17, the T cell receptor gene product expressed by a large fraction of MAM-reactive human T cells.

1991 ◽  
Vol 174 (4) ◽  
pp. 891-900 ◽  
Author(s):  
S M Friedman ◽  
M K Crow ◽  
J R Tumang ◽  
M Tumang ◽  
Y Q Xu ◽  
...  
Keyword(s):  
Monoclonal Antibody ◽  
T Cells ◽  
T Cell ◽  
T Cell Receptor ◽  
Cell Activity ◽  
Cell Receptor ◽  
Cytotoxic T Cell ◽  
Mycoplasma Arthritidis ◽  
Human T Cells

While all known microbial superantigens are mitogenic for human peripheral blood lymphocytes (PBL), the functional response induced by Mycoplasma arthritidis-derived superantigen (MAM) is unique in that MAM stimulation of PBL consistently results in T cell-dependent B cell activation characterized by polyclonal IgM and IgG production. These immunostimulatory effects of MAM on the humoral arm of the human immune system warranted a more precise characterization of MAM-reactive human T cells. Using an uncloned MAM reactive human T cell line as immunogen, we have generated a monoclonal antibody (mAb) (termed C1) specific for the T cell receptor V beta gene expressed by the major fraction of MAM-reactive human T cells, V beta 17. In addition, a V beta 17- MAM-reactive T cell population exists, assessed by MAM, induced T cell proliferation and cytotoxic T cell activity. mAb C1 will be useful in characterizing the functional properties of V beta 17+ T cells and their potential role in autoimmune disease.

scholarly journals Regulatory functions of hapten-reactive helper and suppressor T lymphocytes. II. Selective inactivation of hapten-reactive suppressor T cells by hapten-nonimmunogenic copolymers of D-amino acids, and its application to the study of suppressor T-cell effect on helper T-cell development

1977 ◽  
Vol 146 (1) ◽  
pp. 91-106 ◽  
Author(s):  
T Hamaoka ◽  
M Yoshizawa ◽  
H Yamamoto ◽  
M Kuroki ◽  
M Kitagawa
Keyword(s):  
T Cells ◽  
T Cell ◽  
T Lymphocytes ◽  
T Cell Development ◽  
Cell Activity ◽  
Helper T Cells ◽  
Suppressor T Cells ◽  
Helper T Cell ◽  
Suppressor T Cell ◽  
Suppressor T Lymphocytes

An experimental condition was established in vivo for selectively eliminating hapten-reactive suppressor T-cell activity generated in mice primed with a para-azobenzoate (PAB)-mouse gamma globulin (MGG)-conjugate and treated with PAB-nonimmunogenic copolymer of D-amino acids (D- glutamic acid and D-lysine; D-GL). The elimination of suppressor T-cell activity with PAB-D-GL treatment from the mixed populations of hapten- reactive suppressor and helper T cells substantially increased apparent helper T-cell activity. Moreover, the inhibition of PAB-reactive suppressor T-cell generation by the pretreatment with PAB-D-GL before the PAB-MGG-priming increased the development of PAB-reactive helper T-cell activity. The analysis of hapten-specificity of helper T cells revealed that the reactivity of helper cells developed in the absence of suppressor T cells was more specific for primed PAB-determinants and their cross-reactivities to structurally related determinants such as meta-azobenzoate (MAB) significantly decreased, as compared with the helper T-cell population developed in the presence of suppressor T lymphocytes. In addition, those helper T cells generated in the absence of suppressor T cells were highly susceptible to tolerogenesis by PAB-D- GL. Similarly, the elimination of suppressor T lymphocytes also enhanced helper T-cell activity in a polyclonal fashion in the T-T cell interactions between benzylpenicilloyl (BPO)-reactive T cells and PAB- reactive T cells after immunization of mice with BPO-MGG-PAB. Thus inhibition of BPO-reactive suppressor T-cell development by the BPO-v-GL- pretreatment resulted in augmented generation of PAB-reactive helper T cells with higher susceptibility of tolerogenesis to PAB-D-GL. Thus, these results support the notion that suppressor T cells eventually suppress helper T-cell activity and indicate that the function of suppressor T cells related to helper T-cell development is to inhibit the increase in the specificity and apparent affinity of helper T cells in the primary immune response. The hapten-reactive suppressor and helper T lymphocytes are considered as a model system of T cells that regulate the immune response, and the potential applicability of this system to manipulating various T cell-mediated immune responses is discussed in this context.

scholarly journals Image classification using Deep learning

2018 ◽  
Vol 7 (2.7) ◽  
pp. 614 ◽  
Author(s):  
M Manoj krishna ◽  
M Neelima ◽  
M Harshali ◽  
M Venu Gopala Rao
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Image Processing ◽  
Computer Vision ◽  
Deep Learning ◽  
Image Classification ◽  
Convolutional Neural Networks ◽  
Classical Problem

The image classification is a classical problem of image processing, computer vision and machine learning fields. In this paper we study the image classification using deep learning. We use AlexNet architecture with convolutional neural networks for this purpose. Four test images are selected from the ImageNet database for the classification purpose. We cropped the images for various portion areas and conducted experiments. The results show the effectiveness of deep learning based image classification using AlexNet.  

Sign in / Sign up

Export Citation Format

Share Document