Low False-hit Rates Research Articles

PKT: fast color-based spatial model for human skin detection

We present a new detection method for color-based object detection, which can improve the performance of learning procedures in terms of speed, accuracy, and efficiency, using spatial inference, and algorithm. We applied the model to human skin detection from an image; however, the method can also work for other machine learning tasks involving image pixels. We propose (1) an improved RGB/HSL human skin color threshold to tackle darker human skin color detection problem. (2), we also present a new rule-based fast algorithm (packed k-dimensional tree --- PKT) that depends on an improved spatial structure for human skin/face detection from colored 2D images. We also implemented a novel packed quad-tree (PQT) to speed up the quad-tree performance in terms of indexing. We compared the proposed system to traditional pixel-by-pixel (PBP)/pixel-wise (PW) operation, and quadtree based procedures. The results show that our proposed spatial structure performs better (with a very low false hit rate, very high precision, and accuracy rate) than most state-of-the-art models.

Identification of dual active agents targeting 5-HT1A and SERT by combinatorial virtual screening methods.

Selective serotonin reuptake inhibitors (SSRIs) are most adopted therapeutics marketed for major depression, and the efficacy of which are greatly reduced by their delayed onset of action and undesirable side effects. 5-HT1A receptor partial agonist and SERT inhibitor (SPARI) was proposed as a novel strategy to overcome the shortage of efficacy by a negative feedback control of 5-HT1A receptor. However, only one SPARI (vilazodone) has been approved for clinical use, and none is currently in clinical trial, which demonstrates a strong need for searching more novel SPARIs to facilitate antidepressants discovery. This work applied a combinatorial virtual screening method (CVSM) by integrating multiple tools. Statistic analysis reveals that CVSM surpasses single virtual screening methods in terms of hit rates and enrichment factors. By adopting optimized CVSM, 91 promising dual target leads form 15 scaffolds were identified, and 40% of these scaffolds have already been reported to show antidepressant related therapeutic effects. In sum, CVSM is capable in identifying novel SPARIs from large chemical libraries with extremely low false hit rate.

Hierarchical clustering of EMD based interest points for road sign detection

Abstract This paper presents an automatic road traffic signs detection and recognition system based on hierarchical clustering of interest points and joint transform correlation. The proposed algorithm consists of the three following stages: interest points detection, clustering of those points and similarity search. At the first stage, good discriminative, rotation and scale invariant interest points are selected from the image edges based on the 1-D empirical mode decomposition (EMD). We propose a two-step unsupervised clustering technique, which is adaptive and based on two criterion. In this context, the detected points are initially clustered based on the stable local features related to the brightness and color, which are extracted using Gabor filter. Then points belonging to each partition are reclustered depending on the dispersion of the points in the initial cluster using position feature. This two-step hierarchical clustering yields the possible candidate road signs or the region of interests (ROIs). Finally, a fringe-adjusted joint transform correlation (JTC) technique is used for matching the unknown signs with the existing known reference road signs stored in the database. The presented framework provides a novel way to detect a road sign from the natural scenes and the results demonstrate the efficacy of the proposed technique, which yields a very low false hit rate.

Development and experimental test of support vector machines virtual screening method for searching Src inhibitors from large compound libraries.

BackgroundSrc plays various roles in tumour progression, invasion, metastasis, angiogenesis and survival. It is one of the multiple targets of multi-target kinase inhibitors in clinical uses and trials for the treatment of leukemia and other cancers. These successes and appearances of drug resistance in some patients have raised significant interest and efforts in discovering new Src inhibitors. Various in-silico methods have been used in some of these efforts. It is desirable to explore additional in-silico methods, particularly those capable of searching large compound libraries at high yields and reduced false-hit rates.ResultsWe evaluated support vector machines (SVM) as virtual screening tools for searching Src inhibitors from large compound libraries. SVM trained and tested by 1,703 inhibitors and 63,318 putative non-inhibitors correctly identified 93.53%~ 95.01% inhibitors and 99.81%~ 99.90% non-inhibitors in 5-fold cross validation studies. SVM trained by 1,703 inhibitors reported before 2011 and 63,318 putative non-inhibitors correctly identified 70.45% of the 44 inhibitors reported since 2011, and predicted as inhibitors 44,843 (0.33%) of 13.56M PubChem, 1,496 (0.89%) of 168 K MDDR, and 719 (7.73%) of 9,305 MDDR compounds similar to the known inhibitors.ConclusionsSVM showed comparable yield and reduced false hit rates in searching large compound libraries compared to the similarity-based and other machine-learning VS methods developed from the same set of training compounds and molecular descriptors. We tested three virtual hits of the same novel scaffold from in-house chemical libraries not reported as Src inhibitor, one of which showed moderate activity. SVM may be potentially explored for searching Src inhibitors from large compound libraries at low false-hit rates.

A Two-Step Target Binding and Selectivity Support Vector Machines Approach for Virtual Screening of Dopamine Receptor Subtype-Selective Ligands

Target selective drugs, such as dopamine receptor (DR) subtype selective ligands, are developed for enhanced therapeutics and reduced side effects. In silico methods have been explored for searching DR selective ligands, but encountered difficulties associated with high subtype similarity and ligand structural diversity. Machine learning methods have shown promising potential in searching target selective compounds. Their target selective capability can be further enhanced. In this work, we introduced a new two-step support vector machines target-binding and selectivity screening method for searching DR subtype-selective ligands, which was tested together with three previously-used machine learning methods for searching D1, D2, D3 and D4 selective ligands. It correctly identified 50.6%–88.0% of the 21–408 subtype selective and 71.7%–81.0% of the 39–147 multi-subtype ligands. Its subtype selective ligand identification rates are significantly better than, and its multi-subtype ligand identification rates are comparable to the best rates of the previously used methods. Our method produced low false-hit rates in screening 13.56 M PubChem, 168,016 MDDR and 657,736 ChEMBLdb compounds. Molecular features important for subtype selectivity were extracted by using the recursive feature elimination feature selection method. These features are consistent with literature-reported features. Our method showed similar performance in searching estrogen receptor subtype selective ligands. Our study demonstrated the usefulness of the two-step target binding and selectivity screening method in searching subtype selective ligands from large compound libraries.

A new approach in road sign recognition based on fast fractal coding

The tasks of traffic signs are notifying drivers about the current state of the road and giving them other important information for navigation. In this paper, a new approach for detection, tracking, and recognition such objects is presented. Road signs are detected using color thresholding, after that candidate blobs that have specific criteria are classified based on their geometrical shape and are tracked trough successive frames based on a new similarity measure. Candidate blobs that successfully pass the tracking module are processed for extracting their fractal features, and final recognition is done based on support vector machines with kernel function. Results validate effectiveness of newly employed fractal feature and show high accuracy with a low false hit rate of this method and its robustness to illumination changes and road sign occlusion or scale changes. Also results indicate that compared to the other pictogram feature representation techniques, this approach shows a more proper description of road signs.

Image Segmentation and Shape Analysis for Road-Sign Detection

This paper proposes an automatic road-sign recognition method based on image segmentation and joint transform correlation (JTC) with the integration of shape analysis. The presented system is universal, which is able to detect traffic signs of any countries with any color and any of the existing shapes (e.g., circular, rectangular, triangular, pentagonal, and octagonal) and is invariant to transformation (e.g., translation, rotation, scale, and occlusion). The main contributions of this paper are: 1) the formulation of two new criteria for analyzing different shapes using two basic geometric properties, 2) the recategorization of the rectangular signs into diamond or nondiamond shapes based on the inclination of the four sides with the ground and 3) the employment of the distortion-invariant fringe-adjusted JTC (FJTC) technique for recognition. There are three main stages in the proposed algorithm: 1) segmentation by clustering the pixels based on the color features to find the regions of interest (ROIs); 2) traffic-sign detection by using two novel shape classification criteria, i.e., the relationship between area and perimeter and the number of sides of a given shape; and 3) recognition of the road sign using FJTC to match the unknown signs with the known reference road signs stored in the database. Experimental results on real-life images show a high success rate and a very low false hit rate and demonstrate that the proposed framework is invariant to translation, rotation, scale, and partial occlusions.

Virtual Screening of Selective Multitarget Kinase Inhibitors by Combinatorial Support Vector Machines

Multitarget agents have been increasingly explored for enhancing efficacy and reducing countertarget activities and toxicities. Efficient virtual screening (VS) tools for searching selective multitarget agents are desired. Combinatorial support vector machines (C-SVM) were tested as VS tools for searching dual-inhibitors of 11 combinations of 9 anticancer kinase targets (EGFR, VEGFR, PDGFR, Src, FGFR, Lck, CDK1, CDK2, GSK3). C-SVM trained on 233-1,316 non-dual-inhibitors correctly identified 26.8%-57.3% (majority >36%) of the 56-230 intra-kinase-group dual-inhibitors (equivalent to the 50-70% yields of two independent individual target VS tools), and 12.2% of the 41 inter-kinase-group dual-inhibitors. C-SVM were fairly selective in misidentifying as dual-inhibitors 3.7%-48.1% (majority <20%) of the 233-1,316 non-dual-inhibitors of the same kinase pairs and 0.98%-4.77% of the 3,971-5,180 inhibitors of other kinases. C-SVM produced low false-hit rates in misidentifying as dual-inhibitors 1,746-4,817 (0.013%-0.036%) of the 13.56 M PubChem compounds, 12-175 (0.007%-0.104%) of the 168 K MDDR compounds, and 0-84 (0.0%-2.9%) of the 19,495-38,483 MDDR compounds similar to the known dual-inhibitors. C-SVM was compared to other VS methods Surflex-Dock, DOCK Blaster, kNN and PNN against the same sets of kinase inhibitors and the full set or subset of the 1.02 M Zinc clean-leads data set. C-SVM produced comparable dual-inhibitor yields, slightly better false-hit rates for kinase inhibitors, and significantly lower false-hit rates for the Zinc clean-leads data set. Combinatorial SVM showed promising potential for searching selective multitarget agents against intra-kinase-group kinases without explicit knowledge of multitarget agents.

Identifying Novel Type ZBGs and Nonhydroxamate HDAC Inhibitors Through a SVM Based Virtual Screening Approach.

Histone deacetylase inhibitors (HDACi) have been successfully used for the treatment of cancers and other diseases. Search for novel type ZBGs and development of non-hydroxamate HDACi has become a focus in current research. To complement this, it is desirable to explore a virtual screening (VS) tool capable of identifying different types of potential inhibitors from large compound libraries with high yields and low false-hit rates similar to HTS. This work explored the use of support vector machines (SVM) combined with our newly developed putative non-inhibitor generation method as such a tool. SVM trained by 702 pre-2008 hydroxamate HDACi and 64334 putative non-HDACi showed good yields and low false-hit rates in cross-validation test and independent test using 220 diverse types of HDACi reported since 2008. The SVM hit rates in scanning 13.56 M PubChem and 168K MDDR compounds are comparable to HTS rates. Further structural analysis of SVM virtual hits suggests its potential for identification of non-hydroxamate HDACi. From this analysis, a series of novel ZBG and cap groups were proposed for HDACi design.

Virtual Screening of Abl Inhibitors from Large Compound Libraries by Support Vector Machines

Abl promotes cancers by regulating cell morphogenesis, motility, growth, and survival. Successes of several marketed and clinical trial Abl inhibitors against leukemia and other cancers and appearances of reduced efficacies and drug resistances have led to significant interest in and efforts for developing new Abl inhibitors. In silico methods of pharmacophore, fragment, and molecular docking have been used in some of these efforts. It is desirable to explore other in silico methods capable of searching large compound libraries at high yields and reduced false-hit rates. We evaluated support vector machines (SVM) as a virtual screening tool for searching Abl inhibitors from large compound libraries. SVM trained and tested by 708 inhibitors and 65,494 putative noninhibitors correctly identified 84.4 to 92.3% inhibitors and 99.96 to 99.99% noninhibitors in 5-fold cross validation studies. SVM trained by 708 pre-2008 inhibitors and 65 494 putative noninhibitors correctly identified 50.5% of the 91 inhibitors reported since 2008 and predicted as inhibitors 29,072 (0.21%) of 13.56M PubChem, 659 (0.39%) of 168K MDDR, and 330 (5.0%) of 6638 MDDR compounds similar to the known inhibitors. SVM showed comparable yields and substantially reduced false-hit rates against two similarity based and another machine learning VS methods based on the same training and testing data sets and molecular descriptors. These suggest that SVM is capable of searching Abl inhibitors from large compound libraries at low false-hit rates.

Identification of small molecule aggregators from large compound libraries by support vector machines

Small molecule aggregators non-specifically inhibit multiple unrelated proteins, rendering them therapeutically useless. They frequently appear as false hits and thus need to be eliminated in high-throughput screening campaigns. Computational methods have been explored for identifying aggregators, which have not been tested in screening large compound libraries. We used 1319 aggregators and 128,325 non-aggregators to develop a support vector machines (SVM) aggregator identification model, which was tested by four methods. The first is five fold cross-validation, which showed comparable aggregator and significantly improved non-aggregator identification rates against earlier studies. The second is the independent test of 17 aggregators discovered independently from the training aggregators, 71% of which were correctly identified. The third is retrospective screening of 13M PUBCHEM and 168K MDDR compounds, which predicted 97.9% and 98.7% of the PUBCHEM and MDDR compounds as non-aggregators. The fourth is retrospective screening of 5527 MDDR compounds similar to the known aggregators, 1.14% of which were predicted as aggregators. SVM showed slightly better overall performance against two other machine learning methods based on five fold cross-validation studies of the same settings. Molecular features of aggregation, extracted by a feature selection method, are consistent with published profiles. SVM showed substantial capability in identifying aggregators from large libraries at low false-hit rates.

Evaluation of Virtual Screening Performance of Support Vector Machines Trained by Sparsely Distributed Active Compounds

Virtual screening performance of support vector machines (SVM) depends on the diversity of training active and inactive compounds. While diverse inactive compounds can be routinely generated, the number and diversity of known actives are typically low. We evaluated the performance of SVM trained by sparsely distributed actives in six MDDR biological target classes composed of a high number of known actives (983-1645) of high, intermediate, and low structural diversity (muscarinic M1 receptor agonists, NMDA receptor antagonists, thrombin inhibitors, HIV protease inhibitors, cephalosporins, and renin inhibitors). SVM trained by regularly sparse data sets of 100 actives show improved yields at substantially reduced false-hit rates compared to those of published studies and those of Tanimoto-based similarity searching method based on the same data sets and molecular descriptors. SVM trained by very sparse data sets of 40 actives (2.4%-4.1% of the known actives) predicted 17.5-39.5%, 23.0-48.1%, and 70.2-92.4% of the remaining 943-1605 actives in the high, intermediate, and low diversity classes, respectively, 13.8-68.7% of which are outside the training compound families. SVM predicted 99.97% and 97.1% of the 9.997 M PUBCHEM and 167K remaining MDDR compounds as inactive and 2.6%-8.3% of the 19,495-38,483 MDDR compounds similar to the known actives as active. These suggest that SVM has substantial capability in identifying novel active compounds from sparse active data sets at low false-hit rates.

A bioluminescent assay for monoamine oxidase activity

This article describes a novel two-step homogeneous bioluminescent assay for monoamine oxidase (MAO) that is simple, sensitive, and amenable to high-throughput screening. In the first step, MAO reacts with an aminopropylether analog of methyl ester luciferin. In the second step, a luciferin detection reagent inactivates MAO and converts the product of the first step into a luminescent signal. The amount of light produced is proportional to the amount of MAO and the time of incubation in the first step, but the luminescent signal is stable in the second step with a half-life greater than 5 h. The assay has high precision, is more sensitive than current fluorescent methods, and can accurately measure the binding constants of known substrates and inhibitors. An automated screen of the Sigma–RBI Library of Pharmacologically Active Compounds (LOPAC 1280) revealed a surprisingly high percentage of MAO inhibitors (16%) with a low false hit rate (0.9%). This implies that a significant number of compounds interact with the MAO enzymes and suggests that it is important to include MAO assays in drug metabolism studies. Other advantages of this bioluminescent assay over comparable fluorescent assays are discussed.