DeepCDA: deep cross-domain compound–protein affinity prediction through LSTM and convolutional neural networks

2020 ◽  
Vol 36 (17) ◽  
pp. 4633-4642 ◽  
Author(s):  
Karim Abbasi ◽  
Parvin Razzaghi ◽  
Antti Poso ◽  
Massoud Amanlou ◽  
Jahan B Ghasemi ◽  
...  
Keyword(s):  
Binding Affinity ◽  
Short Term Memory ◽  
Domain Adaptation ◽  
Interaction Strength ◽  
Training Data ◽  
Second Phase ◽  
Cross Domain ◽  
Third Phase ◽  
Reliable Model ◽  
Occurrence Patterns

Abstract Motivation An essential part of drug discovery is the accurate prediction of the binding affinity of new compound–protein pairs. Most of the standard computational methods assume that compounds or proteins of the test data are observed during the training phase. However, in real-world situations, the test and training data are sampled from different domains with different distributions. To cope with this challenge, we propose a deep learning-based approach that consists of three steps. In the first step, the training encoder network learns a novel representation of compounds and proteins. To this end, we combine convolutional layers and long-short-term memory layers so that the occurrence patterns of local substructures through a protein and a compound sequence are learned. Also, to encode the interaction strength of the protein and compound substructures, we propose a two-sided attention mechanism. In the second phase, to deal with the different distributions of the training and test domains, a feature encoder network is learned for the test domain by utilizing an adversarial domain adaptation approach. In the third phase, the learned test encoder network is applied to new compound–protein pairs to predict their binding affinity. Results To evaluate the proposed approach, we applied it to KIBA, Davis and BindingDB datasets. The results show that the proposed method learns a more reliable model for the test domain in more challenging situations. Availability and implementation https://github.com/LBBSoft/DeepCDA.

scholarly journals Distributional Correspondence Indexing for Cross-Lingual and Cross-Domain Sentiment Classification (Extended Abstract)

2018 ◽  
Author(s):  
Alejandro Moreo Fernández ◽  
Andrea Esuli ◽  
Fabrizio Sebastiani
Keyword(s):  
Domain Adaptation ◽  
State Of The Art ◽  
Sentiment Classification ◽  
Training Data ◽  
Target Domain ◽  
Source Domain ◽  
Machine Learning Methods ◽  
Cross Domain ◽  
Current State ◽  
Cross Lingual

Domain Adaptation (DA) techniques aim at enabling machine learning methods learn effective classifiers for a “target” domain when the only available training data belongs to a different “source” domain. In this extended abstract, we briefly describe our new DA method called Distributional Correspondence Indexing (DCI) for sentiment classification. DCI derives term representations in a vector space common to both domains where each dimension reflects its distributional correspondence to a pivot, i.e., to a highly predictive term that behaves similarly across domains. The experiments we have conducted show that DCI obtains better performance than current state-of-the-art techniques for cross-lingual and cross-domain sentiment classification.

scholarly journals Distributional Correspondence Indexing for Cross-Lingual and Cross-Domain Sentiment Classification.

2016 ◽  
Vol 55 ◽  
pp. 131-163 ◽  
Author(s):  
Alejandro Moreo Fernández ◽  
Andrea Esuli ◽  
Fabrizio Sebastiani
Keyword(s):  
Domain Adaptation ◽  
State Of The Art ◽  
Computational Cost ◽  
Sentiment Classification ◽  
Training Data ◽  
Target Domain ◽  
Machine Learning Methods ◽  
Cross Domain ◽  
Current State ◽  
Cross Lingual

Domain Adaptation (DA) techniques aim at enabling machine learning methods learn effective classifiers for a "target'' domain when the only available training data belongs to a different "source'' domain. In this paper we present the Distributional Correspondence Indexing (DCI) method for domain adaptation in sentiment classification. DCI derives term representations in a vector space common to both domains where each dimension reflects its distributional correspondence to a pivot, i.e., to a highly predictive term that behaves similarly across domains. Term correspondence is quantified by means of a distributional correspondence function (DCF). We propose a number of efficient DCFs that are motivated by the distributional hypothesis, i.e., the hypothesis according to which terms with similar meaning tend to have similar distributions in text. Experiments show that DCI obtains better performance than current state-of-the-art techniques for cross-lingual and cross-domain sentiment classification. DCI also brings about a significantly reduced computational cost, and requires a smaller amount of human intervention. As a final contribution, we discuss a more challenging formulation of the domain adaptation problem, in which both the cross-domain and cross-lingual dimensions are tackled simultaneously.

scholarly journals Fault Diagnosis of Gearbox with a Transferable Deep Neural Network

2020 ◽  
Author(s):  
Greg Smith ◽  
Masayoshi Shibatani
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Deep Learning ◽  
Fault Diagnosis ◽  
Domain Adaptation ◽  
Fault Detection And Diagnosis ◽  
Training Data ◽  
Operating Condition ◽  
Cross Domain ◽  
Testing Data

In the past years, various intelligent machine learning and deep learning algorithms have been developed and widely applied for gearbox fault detection and diagnosis. However, the real-time application of these intelligent algorithms has been limited, mainly due to the fact that the model developed using data from one machine or one operating condition has serious diagnosis performance degradation when applied to another machine or the same machine with a different operating condition. The reason for poor model generalization is the distribution discrepancy between the training and testing data. This paper proposes to address this issue using a deep learning based cross domain adaptation approach for gearbox fault diagnosis. Labelled data from training dataset and unlabeled data from testing dataset is used to achieve the cross-domain adaptation task. A deep convolutional neural network (CNN) is used as the main architecture. Maximum mean discrepancy is used as a measure to minimize the distribution distance between the labelled training data and unlabeled testing data. The study proposes to reduce the discrepancy between the two domains in multiple layers of the designed CNN to adapt the learned representations from the training data to be applied in the testing data. The proposed approach is evaluated using experimental data from a gearbox under significant speed variation and multiple health conditions. An appropriate benchmarking with both traditional machine learning methods and other domain adaptation methods demonstrates the superiority of the proposed method.

TrCSVM: a novel approach for the classification of melanoma skin cancer using transfer learning

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Lokesh Singh ◽  
Rekh Ram Janghel ◽  
Satya Prakash Sahu
Keyword(s):  
Support Vector Machine ◽  
Skin Lesion ◽  
Transfer Learning ◽  
Domain Adaptation ◽  
Training Data ◽  
Support Vector ◽  
Second Phase ◽  
Target Domain ◽  
Content Type

PurposeThe study aims to cope with the problems confronted in the skin lesion datasets with less training data toward the classification of melanoma. The vital, challenging issue is the insufficiency of training data that occurred while classifying the lesions as melanoma and non-melanoma.Design/methodology/approachIn this work, a transfer learning (TL) framework Transfer Constituent Support Vector Machine (TrCSVM) is designed for melanoma classification based on feature-based domain adaptation (FBDA) leveraging the support vector machine (SVM) and Transfer AdaBoost (TrAdaBoost). The working of the framework is twofold: at first, SVM is utilized for domain adaptation for learning much transferrable representation between source and target domain. In the first phase, for homogeneous domain adaptation, it augments features by transforming the data from source and target (different but related) domains in a shared-subspace. In the second phase, for heterogeneous domain adaptation, it leverages knowledge by augmenting features from source to target (different and not related) domains to a shared-subspace. Second, TrAdaBoost is utilized to adjust the weights of wrongly classified data in the newly generated source and target datasets.FindingsThe experimental results empirically prove the superiority of TrCSVM than the state-of-the-art TL methods on less-sized datasets with an accuracy of 98.82%.Originality/valueExperiments are conducted on six skin lesion datasets and performance is compared based on accuracy, precision, sensitivity, and specificity. The effectiveness of TrCSVM is evaluated on ten other datasets towards testing its generalizing behavior. Its performance is also compared with two existing TL frameworks (TrResampling, TrAdaBoost) for the classification of melanoma.

PENERAPAN MODEL QUANTUM LEARNING UNTUK MENINGKATKAN HASIL BELAJAR AUTOCAD SISWA KELAS X TEKNIK PEMESINAN SMK NEGERI 1 STABAT

2013 ◽  
Vol 5 (1) ◽  
Author(s):  
Abdul Hasan Saragih
Keyword(s):  
Positive Response ◽  
Phase 1 ◽  
First Grade ◽  
Learning Model ◽  
Second Phase ◽  
Phase 2 ◽  
Phase 3 ◽  
The Third ◽  
Third Phase ◽  
Quantum Learning

This classroom research was conducted on the autocad instructions to the first grade of mechinary class of SMK Negeri 1 Stabat aiming at : (1) improving the student’ archievementon autocad instructional to the student of mechinary architecture class of SMK Negeri 1 Stabat, (2) applying Quantum Learning Model to the students of mechinary class of SMK Negeri 1 Stabat, arising the positive response to autocad subject by applying Quantum Learning Model of the students of mechinary class of SMK Negeri 1 Stabat. The result shows that (1) by applying quantum learning model, the students’ achievement improves significantly. The improvement ofthe achievement of the 34 students is very satisfactory; on the first phase, 27 students passed (70.59%), 10 students failed (29.41%). On the second phase 27 students (79.41%) passed and 7 students (20.59%) failed. On the third phase 30 students (88.24%) passed and 4 students (11.76%) failed. The application of quantum learning model in SMK Negeri 1 Stabat proved satisfying. This was visible from the activeness of the students from phase 1 to 3. The activeness average of the students was 74.31% on phase 1,81.35% on phase 2, and 83.63% on phase 3. (3) The application of the quantum learning model on teaching autocad was very positively welcome by the students of mechinary class of SMK Negeri 1 Stabat. On phase 1 the improvement was 81.53% . It improved to 86.15% on phase 3. Therefore, The improvement ofstudent’ response can be categorized good.

Mixed-Methods Evaluation as Scaffold to the Development of a Mobile Health Application to Support the Prescription of Wheelchairs and Postural Support Devices (Preprint)

2020 ◽  
Author(s):  
Adriana Klein ◽  
Roseli de Deus Lopes ◽  
Rodrigo Suigh
Keyword(s):  
Mixed Methods ◽  
Mobile Health ◽  
Second Phase ◽  
The Third ◽  
Evaluation Approach ◽  
Mixed Methods Evaluation ◽  
Usability Tests ◽  
Third Phase ◽  
Postural Support ◽  
Support Devices

BACKGROUND EasySeating is a mobile health (mHealth) app that supports the prescription of wheelchair and postural support devices (WPSD). It can be used by occupational therapists (OT) and physiotherapists (PT) who prescribe WPSD. The app offers a standardization of the prescription procedure, showing images, metrics and details that guide the prescriber to decide on the best equipment. It was developed with an iterative mixed-methods evaluation approach. Objective: The aim of this study was to investigate the processes involved in the prescription of WPSD and to propose, develop and evaluate a mHealth to support OT and PT prescribers. OBJECTIVE The aim of this study was to investigate the processes involved in the prescription of WPSD and to propose, develop and evaluate a mHealth to support OT and PT prescribers. METHODS This study was divided into three phases and was carried out as an iterative process composed of user consulting/testing (using a mixed-methods evaluation approach), system (re)design and software development. The first phase consisted of the collection of qualitative and quantitative data to map and understand the users requirements and of the development of the first prototype (v1) of the app. This data collection was performed through semi-structured interviews with 14 OT and PT prescribers, 5 specialized technicians and 5 WPSD users. The second phase aimed at improving the overall functionality of the app and consisted in the development, test and evaluation of the prototypes v1, v2, v3 and v4. A total of 59 prescribers tested and evaluated these prototypes by means of open interviews, semi-structured questionnaires and focus groups. The third phase focused in the usability aspects of the app. It consisted in the development and test of the prototype v5. Eight technology specialists assessed its usability through heuristics evaluation. RESULTS Data collected in phase one indicated there is a lack of standardization on the prescription of postural support devices (PSD). A divergent nomenclature for the PSDs was also found and classified in eight categories. These information guided the development of the first prototype of the EasySeating app. Phase two results pointed that the prescribers value the insertion of the app into their clinical practice, as it accelerates and increases the quality of the evaluation process and improves the organization of the prescription information. Significant suggestions for the improvement of the app were given during the users tests, including the use of images to represent the PSDs. The usability tests from the third phase revealed two strong issues that must be solved: the need of greater feedback and failures in the persistence of the input data. CONCLUSIONS This study demonstrated that there is a lack of systematization of the WPSD prescription process. The evaluation of the developed EasySeating app demonstrated that there is a potential to standardize, integrate and organize the WPSD prescription information, supporting and facilitating the decision making process of the prescribers. CLINICALTRIAL This study was approved by the Research Ethics Board of the Universidade de São Paulo (registered protocol n°53929516.6.0000.0065) URL - http://plataformabrasil.saude.gov.br/login.jsf

Foucault as Superauthor

2019 ◽  
Vol 29 (2) ◽  
pp. 104-125
Keyword(s):  
Free Speech ◽  
Second Phase ◽  
Contemporary Society ◽  
Sovereign Power ◽  
Ancient Greek ◽  
Main Motivation ◽  
Critical Attitude ◽  
The Third ◽  
Third Phase ◽  
Lecture Courses

Three phases in Foucault’s examination of authorship and free speech were essential to him throughout his life. They can be linked to such texts as the three lectures “What is an Author?” (first phase), “What is Critique?,” and “What is Revolution?” (second phase), and the two lecture courses, “Fearless Speech,” and “The Courage of Truth” (third phase). Initially, Foucault merely describes the founders of discursivity (hence, “superauthors”), among whom he reckoned only Marx and Freud, as the sole alternative to his own conceptualization of the author function, which is exhibited en masse in contemporary society. He then modifies his views on superauthorship by making Kant the paradigm and by linking his own concept of free speech to a Kan-tian critical attitude. However, Foucault claims only the half of Kant’s philosophical legacy that is related to the study of the ontology of the self.The article advances the hypothesis that the sovereign power of speech, which can be found in Marx and Heidegger and in generally in the concept of “superauthorship,” becomes unacceptable for Foucault. During the third phase, the danger of a tyrannical use of free speech compels Foucault to make a number of fruitful but questionable choices in his work. He focuses on a single aspect of free speech in which a speaker is in a weaker position and therefore has to overcome his fear in order to tell the truth. Foucault associates this kind of free speech with the ancient Greek notion of parrhesia, which according to his interpretation means “fearless speech”; however, this reading is not always supported by the ancient Greek sources. Foucault’s deliberations bring him to the radical conclusion that free speech transforms into performative “aesthetics of existence.” Foucault’s main motivation for pursuing this line of thought all through his life was to investigate his own abilities and powers as an author

scholarly journals Multi-Path and Group-Loss-Based Network for Speech Emotion Recognition in Multi-Domain Datasets

Sensors ◽  
2021 ◽  
Vol 21 (5) ◽  
pp. 1579 ◽  
Author(s):  
Kyoung Ju Noh ◽  
Chi Yoon Jeong ◽  
Jiyoun Lim ◽  
Seungeun Chung ◽  
Gague Kim ◽  
...  
Keyword(s):  
Emotion Recognition ◽  
Short Term Memory ◽  
Domain Adaptation ◽  
Classification Model ◽  
Speech Emotion Recognition ◽  
Target Domain ◽  
Model Generalization ◽  
Speech Database ◽  
Emotion Labels ◽  
Temporal Feature

Speech emotion recognition (SER) is a natural method of recognizing individual emotions in everyday life. To distribute SER models to real-world applications, some key challenges must be overcome, such as the lack of datasets tagged with emotion labels and the weak generalization of the SER model for an unseen target domain. This study proposes a multi-path and group-loss-based network (MPGLN) for SER to support multi-domain adaptation. The proposed model includes a bidirectional long short-term memory-based temporal feature generator and a transferred feature extractor from the pre-trained VGG-like audio classification model (VGGish), and it learns simultaneously based on multiple losses according to the association of emotion labels in the discrete and dimensional models. For the evaluation of the MPGLN SER as applied to multi-cultural domain datasets, the Korean Emotional Speech Database (KESD), including KESDy18 and KESDy19, is constructed, and the English-speaking Interactive Emotional Dyadic Motion Capture database (IEMOCAP) is used. The evaluation of multi-domain adaptation and domain generalization showed 3.7% and 3.5% improvements, respectively, of the F1 score when comparing the performance of MPGLN SER with a baseline SER model that uses a temporal feature generator. We show that the MPGLN SER efficiently supports multi-domain adaptation and reinforces model generalization.

scholarly journals Structural and Thermodynamic Analysis of the Resistance Development to Pimodivir (VX-787), the Clinical Inhibitor of Cap Binding to PB2 Subunit of Influenza A Polymerase

Molecules ◽  
2021 ◽  
Vol 26 (4) ◽  
pp. 1007
Author(s):  
Jiří Gregor ◽  
Kateřina Radilová ◽  
Jiří Brynda ◽  
Jindřich Fanfrlík ◽  
Jan Konvalinka ◽  
...  
Keyword(s):  
Thermodynamic Analysis ◽  
Influenza A ◽  
Polymerase Activity ◽  
Second Phase ◽  
Missense Mutations ◽  
Wild Type ◽  
Resistance Development ◽  
Third Phase ◽  
Major Mutation ◽  
Control And Prevention

Influenza A virus (IAV) encodes a polymerase composed of three subunits: PA, with endonuclease activity, PB1 with polymerase activity and PB2 with host RNA five-prime cap binding site. Their cooperation and stepwise activation include a process called cap-snatching, which is a crucial step in the IAV life cycle. Reproduction of IAV can be blocked by disrupting the interaction between the PB2 domain and the five-prime cap. An inhibitor of this interaction called pimodivir (VX-787) recently entered the third phase of clinical trial; however, several mutations in PB2 that cause resistance to pimodivir were observed. First major mutation, F404Y, causing resistance was identified during preclinical testing, next the mutation M431I was identified in patients during the second phase of clinical trials. The mutation H357N was identified during testing of IAV strains at Centers for Disease Control and Prevention. We set out to provide a structural and thermodynamic analysis of the interactions between cap-binding domain of PB2 wild-type and PB2 variants bearing these mutations and pimodivir. Here we present four crystal structures of PB2-WT, PB2-F404Y, PB2-M431I and PB2-H357N in complex with pimodivir. We have thermodynamically analysed all PB2 variants and proposed the effect of these mutations on thermodynamic parameters of these interactions and pimodivir resistance development. These data will contribute to understanding the effect of these missense mutations to the resistance development and help to design next generation inhibitors.

Sign in / Sign up

Export Citation Format

Share Document