ISSN (Print): 1570-1786
ISSN (Online): 1875-6255
Volume 18, 12 Issues, 2021
Download PDF Flyer
Open Access Funding
Promote Your Article
ISSN (Print): 1570-1786
ISSN (Online): 1875-6255
Aims & Scope
Chemical Abstracts Service/SciFinder, Chemistry Citation Index®, ChemWeb, CNKI Scholar, Current Contents®/Physical, Chemical & Earth Sciences, Dimensions, EBSCO, ERA 2018, Genamics JournalSeek, Google Scholar, InCites, Index Copernicus, ISI Alerting Service, J-Gate, Journal Citation Reports/Science Edition, JournalTOCs, MediaFinder®-Standard Periodical Directory, PubsHub, QOAM, Reaxys, Science Citation Index Expanded™ (SciSearch®), Scilit, Scopus, Suweco CZ, TOC Premier, and Ulrich's Periodicals Directory.
Ranking and Category:
Submit Manuscripts Online
Animated Abstract Submission
View Full Editorial Board
5 - Year: 0.865
Self Archiving Policies
Instructions for Authors
Free Copies Online
Open Access Articles
Most Cited Articles
Advertise With Us
Most Accessed Articles
Most Popular Articles
Special Issue Submission
"Letters in Organic Chemistry offers a high impact vehicle for reporting exciting new research."
Univ. of Alberta, Canada
Development and application of feature selection techniques in protein data analysis and prediction
Letters in Organic Chemistry, Volume 14, Number 9
Guest Editor(s): Hao Lin
"It was a nice and delightful experience to work and publish our work with the Bentham Science Publishers. Your cooperation and prompt decisions are highly appreciated during the whole process from submission to final publication. I had a very good experience with you"
Md Yousuf Ansari (M.M. College of Pharmacy, Maharishi Markandeshwar University, Mullana, Ambala, Haryana 133207, India)
With the development of high-throughput sequencing techniques, more and more biological data is available, including
DNA, RNA and protein. Using traditional biochemical method to identify and analyze these biological organic molecule becomes
more difficult because of its expensive experimental materials and long experimental period. To overcome the disadvantage,
computational methods are a good choice. Thus, it is more popular to develop various computational methods to discriminate
molecular types, predict molecular function and identify drug targets because these techniques can extract the essential
characteristics of research object and improve accuracies of models, which is needed by all biological scholars. This special
issue focused on various aspects of the development and application of machine learning techniques in biological organic molecular
In the past forty years, the line of reverse biology (sequence-structure-function) guides theoretical biologists to do researches.
In this special issue, thirteen works were published. Five works designed powerful machine learning methods to predict
special protein function. Feng and Xie  developed a support vector machine-based method to identify malaria mitochondrial
proteins by using sequence information and predicted secondary structure. Their final model could produce the overall
accuracy of >97% in cross-validation test. Yang et al.  proposed a model to identify phage virion proteins and achieved the
accuracy of 97.40% in 10-fold cross-validation. Akbar et al.  introduced three kinds of feature descriptors to formulate antifreeze
proteins (AFP) for AFP prediction. They obtained the accuracy of 95.02% in 10-fold cross-validation. Wang et al. 
presented a predicted model to discriminate between acidic and alkaline enzymes. Their model could also produce encouraging
results. Wang et al.  focused on the autophage-raleted modules in cancer and revealed the potential relationships between
autophagy and colorectal cancer.
Protein post-translation modification (PTM) site prediction is a very hot topic in the field of bioinformatics. Thus, the thematic
issue has accepted and published four papers which focused on protein PTM site identification by using computational
methods. Zhang et al.  used KNN algorithm with BLOSUM matrix to predict phosphorylation site. Their method could produce
the average accuracy of 88.74% for three kinds of residues. Khan et al.  focused on S-nitrosylation prediction and obtained
good results by using neural network. Two kinds of PTMs that are acetylation and succinylation were studied by Xu et
al. . They designed a multilabel learning to the topic and achieved encouraging results. Another paper  proposed a Deep
Sparse Auto-encoder to recognize A-to-I site. And a free webserver was established for users.
Generally, protein functions were determined by their structure. There are three papers in this issue which are focused on
protein structure and interaction. Kong et al.  proposed the use of predicted secondary structure to classify four kinds of
protein structural classes. Very high accuracies were achieved, suggesting their model’s power. Wu et al.  combined
mRMR with SVM to establish QSAR model. In jackknife cross-validation, their model could achieve the mean relative errors
of 1.72%. Protein-protein interaction (PPI) is very important in biological process studies. Zhang et al.  utilized weighted
feature fusion to predict PPI. Their proposed method achieved 99.59% sensitivity and 93.66% prediction accuracy.
The thirteenth paper  focused on the prediction of epidemic avian influenza according to goodle information. The model
will provide convenience to the users in hospital and health providers.
In summary, computational methods have been widely applied in variance of biological and chemical fields. These techniques
can fast and accurately identify functional molecule, extract the essential characteristics of research object and improve
accuracies of models, which are needed by all scholars.
No Text Found