Search

The risk of emerging infectious diseases (EID) is increasing globally. More than 60% of EIDs worldwide are caused by animal-borne pathogens, and most viral pathogens are rodent-borne. This study aimed to characterise the virome and analyse the phylogenetic evolution and diversity of rodent-borne viruses in Hainan Province, China. We collected 588 anal and throat samples from rodents, combined them into 28 pools according to their species and location, and processed them for next-generation sequencing and bioinformatics analysis. The diverse viral reads closely related to mammals were assigned to 15 viral families. Molecular clues of the important rodent-borne viruses were further identified by polymerase chain reaction for phylogenetic analysis and annotation of genetic characteristics such as coronavirus, arenavirus, picornavirus. We identified a pestivirus in Leopoldoms edwardsi and two bocaviruses in Rattus andamanensis and Leopoldoms edwardsi from the national nature reserves of Jianfengling and Bangxi with low amino acid identity to known pathogens are proposed as the novel species, and their rodent hosts have not been previously reported to carry these viruses. These results expand our knowledge of viral classification and host range and suggest that there are highly diverse, undiscovered viruses that have evolved independently in their unique wildlife hosts in inaccessible areas, which may cause zoonosis if they cross their host barrier. Our virome and phylogenetic analyses of rodent-borne viruses provide basic data for the prevention and control of human infectious diseases caused by rodent-borne viruses in the subtropical area of China.

Dengue virus is an arthropod-borne pathogen that is transmitted to humans primarily by Aedes spp. mosquitos, causing the acute infectious disease, dengue fever (DF). Until 2019, no dengue outbreak had been reported in Hainan Province for over 20 years. However, in early September of 2019, an increasing number of infected cases appeared and the DF outbreak lasted for over one month in Haikou City, Hainan Province. In our study, we collected 97 plasma samples from DF patients at three hospitals, as well as 1585 mosquito larvae samples from puddles in different areas of Haikou. There were 49 (50.5%) plasma samples found to be strongly positive and 9 (9.3%) plasma samples were weakly positive against the NS1 antigen. We discovered DENV both in the patient’s plasma samples and mosquito larvae samples, and isolated the virus from C6/36 cells inoculated with the acute phase serum of patients. Phylogenetic analysis revealed that the new strains were the most closely related to the epidemic strain in the southern regions of China, belonging to lineage IV, genotype I, DENV-1. Compared to the seven closest strains from neighboring countries and provinces, a total of 18 amino acid mutations occurred in the coding sequences (CDS) of the new isolated strain, DENV1 HMU-HKU-2. Our data shows that dengue virus is re-emerged in Hainan, and pose new threats for public health. Thus regular molecular epidemiological surveillance is necessary for control and prevention of DENV transmission. Electronic supplementary material The online version of this article (10.1007/s12250-020-00335-x) contains supplementary material, which is available to authorized users.

Astroviruses infect human and animals and cause diarrhea, fever, and vomiting. In severe cases, these infections may be fatal in infants and juvenile animals. Previous evidence showed that humans in contact with infected animals can develop serological responses to astroviruses. Mamastrovirus 11 is a species of Mamastrovirus and was first reported in 2018. It was detected in the fecal samples of a California sea lion. The genome sequence of its capsid protein (CP) was submitted to GenBank. However, the genome sequence of its non-structural protein region was not elucidated. In the present study, we characterized the genome sequences of the novel astroviruses AstroV-HMU-1 and AstroV-like-HMU-2. These were obtained from California sea lions (Zalophus californianus) and walruses (Odobenus rosmarus) presenting with loose stools. A phylogenetic analysis revealed that the CP of AstroV-HMU-1 closely clustered with Mamastrovirus 11 while its RNA-dependent RNA polymerase (RdRp) and serine protease (SP) were closely related to the mink astrovirus in the genus Mamastrovirus. The genome of AstroV-HMU-1 provided basic information regarding the NS protein regions of Mamastrovirus 11. Recombination analyses showed that the genomes of Z. californianus AstroV-HMU-1, VA2/human and the mink astrovirus may have recombined long ago. The NS of AstroV-like-HMU-2 segregated from the Astroviridae in the deep root of the phylogenetic tree and exhibited 36% amino acid identity with other mamastroviruses. Thus, AstroV-like-HMU-2 was proposed as a member of a new genus in the unclassified Astroviridae. The present study suggested that that the loose stools of pinnipeds may be the result of occasional infection by this novel astrovirus. This discovery provides a scientific basis for future investigations into other animal-borne infectious diseases.

Scrub typhus, caused by mite-borne Orientia tsutsugamushi (O. tsutsugamushi), is a major febrile disease in the Asia-Pacific region. The DNA load of O. tsutsugamushi in the blood was previously found to be significantly higher in patients with fatal disease than those with non-fatal disease and correlated with the duration of illness, presence of eschar, and hepatic enzyme levels. In this prospective observation study, we analyzed the association of bacterial DNA load with clinical features, disease severity, and genotype using real-time PCR targeting the 56 kDa TSA gene of O. tsutsugamushi in the blood samples of 117 surviving patients with scrub typhus who had not received appropriate antibiotic treatment. The median O. tsutsugamushi DNA load was 3.11×103 copies/mL (range, 44 to 3.3×106 copies/mL). The severity of patients was categorized as mild, moderate, and severe based on the number of dysfunctional organs, and no significant difference in O. tsutsugamushi DNA load was found among these groups. Patients infected with the Karp group showed a significantly higher O. tsutsugamushi DNA load than those in the Gilliam (P P O. tsutsugamushi DNA load (ρ = 0.272, P = 0.022). Correlation analyses indicated that the serum total bilirubin level was positively correlated with O. tsutsugamushi DNA load. In conclusion, the findings in this study demonstrated the association of DNA load of O. tsutsugamushi with the severity and genotype in patients with scrub typhus in Hainan, China.

Currently, multiplex-PCR with genotype-specific primers is widely used for preliminary screening of hepatitis B virus (HBV) genotyping, despite its relatively lower accuracy compared with whole genome sequencing. Here, we present the discrepant results of HBV genotyping by PCR and full-length sequencing. HBV DNA was isolated from chronic hepatitis B serum and the HBV genotype was detected by PCR using genotype-specific primers and full-length genome sequencing. As a result, the determination of genotype B by the PCR method was consistent with the DNA sequencing results; however, PCR revealed that genotype C exhibited a mixed genotype of B and C in the current study. In conclusion, the PCR-based genotyping method may not provide accurate information of the HBV genotype and whole genome sequencing remains the “gold standard” method for HBV genotyping.

The discovery of new viruses is important for predicting their potential threats to the health of humans and other animals. A novel picornavirus was identified from oral, throat, and anal swab samples collected from belugas (Delphinapterus leucas), from Dalian Sun Asia Tourism Holding Co., China, between January and December 2018, using a metagenomics approach. The genome of this novel PicoV-HMU-1 strain was 8197 nucleotides (nt) in length, with a open reading frame (from 1091 to 8074 nt) that encoded a polyprotein precursor of 2328 amino acids. Moreover, the genomic length and GC content of PicoV-HMU-1 were within the ranges found in other picornaviruses, and the genome organization was also similar. Nevertheless, PicoV-HMU-1 had a lower amino acid identity and distinct host species compared with other members of the Picornaviridae family. Phylogenetic trees were constructed based on the P1 and 3D amino acid sequences of PicoV-HMU-1 along with representative members of the Picornaviridae family, which showed that PicoV-HMU-1 was related to unclassified bat picornaviruses groups. These findings suggest that the PicoV-HMU-1 strain represents a potentially novel genus of picornavirus. These data can enhance our understanding of the picornavirus genetic diversity and evolution.

Appropriate timing of cervical remodeling (CR) is key to normal term parturition. To date, mechanisms behind normal and abnormal (premature or delayed) CR remain unclear. Recent studies show regional differences exist in human cervical tissue structure. While the entire cervix contains extracellular matrix (ECM), the internal os is highly cellular containing 50-60% cervical smooth muscle (CSM). The external os contains 10-20% CSM. Previously, we reported ECM rigidity and different ECM proteins influence CSM cell function, highlighting the importance of understanding not only how cervical cells orchestrate cervical ECM remodeling in pregnancy, but also how changes in specific ECM proteins can influence resident cellular function. To understand this dynamic process, we utilized a systematic proteomic approach to understand which soluble ECM and cellular proteins exist in the different regions of the human cervix and how the proteomic profiles change from the non-pregnant (NP) to the pregnant (PG) state. We found the human cervix proteome contains at least 4548 proteins and establish the types and relative abundance of cellular and soluble matrisome proteins found in the NP and PG human cervix. Further, we report the relative abundance of proteins involved with elastic fiber formation and ECM organization/degradation were significantly increased while proteins involved in RNA polymerase I/promoter opening, DNA methylation, senescence, immune system, and compliment activation were decreased in the PG compared to NP cervix. These findings establish an initial platform from which we can further comprehend how changes in the human cervix proteome results in normal and abnormal CR.

Background Inner ear diagnostics is limited by the inability to atraumatically obtain samples of inner ear fluid. The round window membrane (RWM) is an attractive portal for accessing perilymph samples as it has been shown to heal within one week after the introduction of microperforations. A 1 µL volume of perilymph is adequate for proteome analysis, yet the total volume of perilymph within the scala tympani of the guinea pig is limited to less than 5 µL. This study investigates the safety and reliability of a novel hollow microneedle device to aspirate perilymph samples adequate for proteomic analysis. Methods The guinea pig RWM was accessed via a postauricular surgical approach. 3D-printed hollow microneedles with an outer diameter of 100 µm and an inner diameter of 35 µm were used to perforate the RWM and aspirate 1 µL of perilymph. Two perilymph samples were analyzed by liquid chromatography-mass spectrometry-based quantitative proteomics as part of a preliminary study. Hearing was assessed before and after aspiration using compound action potential (CAP) and distortion product otoacoustic emissions (DPOAE). RWMs were harvested 72 h after aspiration and evaluated for healing using confocal microscopy. Results There was no permanent damage to hearing at 72 h after perforation as assessed by CAP (n = 7) and DPOAE (n = 8), and all perforations healed completely within 72 h (n = 8). In the two samples of perilymph analyzed, 620 proteins were detected, including the inner ear protein cochlin, widely recognized as a perilymph marker. Conclusion Hollow microneedles can facilitate aspiration of perilymph across the RWM at a quality and volume adequate for proteomic analysis without causing permanent anatomic or physiologic dysfunction. Microneedles can mediate safe and effective intracochlear sampling and show great promise for inner ear diagnostics.

In vitro protein stability studies are commonly conducted via thermal or chemical denaturation/renaturation of protein. Conventional data analyses on the protein unfolding/(re)folding require well-defined pre- and post-transition baselines to evaluate Gibbs free-energy change associated with the protein unfolding/(re)folding. This evaluation becomes problematic when there is insufficient data for determining the pre- or post-transition baselines. In this study, fitting on such partial data obtained in protein chemical denaturation is established by introducing second-order differential (SOD) analysis to overcome the limitations that the conventional fitting method has. By reducing numbers of the baseline-related fitting parameters, the SOD analysis can successfully fit incomplete chemical denaturation data sets with high agreement to the conventional evaluation on the equivalent completed data, where the conventional fitting fails in analyzing them. This SOD fitting for the abbreviated isothermal chemical denaturation further fulfills data analysis methods on the insufficient data sets conducted in the two prevalent protein stability studies.

Guiding assembly system based on AR is to apply the technol-ogy of Augmented Reality to guide mechanical assembly op-eration and teach or train freshmen. By the means of adapting AR glasses(HoloLens, Magic Leap One etc.), we use virtual information to guide operating real mechanic. AR operation can achieve better effect than traditional reality training, and make industrial assembly process easier for costumers to un-derstand. It can also decrease losses caused by operational error, therefore increases efficiency safety and productivity.

To improve the ability to move from preclinical trials in mouse models of Huntington's disease (HD) to clinical trials in humans, biomarkers are needed that can track similar aspects of disease progression across species. Brain metabolites, detectable by magnetic resonance spectroscopy (MRS), have been suggested as potential biomarkers in HD. In this study, the R6/2 transgenic mouse model of HD was used to investigate the relative sensitivity of the metabolite profiling and the brain volumetry to anticipate the disease progression. Magnetic resonance imaging (MRI) and 1H MRS data were acquired at 9.4 T from the R6/2 mice and wild-type littermates at 4, 8, 12, and 15 weeks. Brain shrinkage was detectable in striatum, cortex, thalamus, and hypothalamus by 12 weeks. Metabolite changes in cortex paralleled and sometimes preceded those in striatum. The entire set of metabolite changes was compressed into principal components (PCs) using Partial Least Squares-Discriminant Analysis (PLS-DA) to increase the sensitivity for monitoring disease progression. In comparing the efficacy of volume and metabolite measurements, the cortical PC1 emerged as the most sensitive single biomarker, distinguishing R6/2 mice from littermates at all time points. Thus, neurochemical changes precede volume shrinkage and become potential biomarkers for HD mouse models.

BACKGROUND/AIM: Esophageal cancer (EC) is a common malignancy with significant morbidity and mortality. As individual cancers exhibit unique mutation patterns, identifying and characterizing gene mutations in EC that may serve as biomarkers might help predict patient outcome and guide treatment. Traditionally, personalized cancer DNA sequencing was impractical and expensive. Recent technological advancements have made targeted DNA sequencing more cost- and time-effective with reliable results. This technology may be useful for clinicians to direct patient treatment. MATERIALS AND METHODS: The Ion PGM and AmpliSeq Cancer Panel was used to identify mutations at 737 hotspot loci of 45 cancer-related genes in 64 EC samples from Chinese patients. RESULTS: Frequent mutations were found in TP53 and less frequent mutations in PIK3CA, FBXW7 and KRAS. CONCLUSION: These results demonstrate that targeted sequencing can reliably identify mutations in individual tumors that make this technology a possibility for clinical use.

In vitro protein stability studies are commonly conducted via thermal or chemical denaturation/renaturation of protein. Conventional data analyses on the protein unfolding/(re)folding require well‐defined pre‐ and post‐transition baselines to evaluate Gibbs free‐energy change associated with the protein unfolding/(re)folding. This evaluation becomes problematic when there is insufficient data for determining the pre‐ or post‐transition baselines. In this study, fitting on such partial data obtained in protein chemical denaturation is established by introducing second‐order differential (SOD) analysis to overcome the limitations that the conventional fitting method has. By reducing numbers of the baseline‐related fitting parameters, the SOD analysis can successfully fit incomplete chemical denaturation data sets with high agreement to the conventional evaluation on the equivalent completed data, where the conventional fitting fails in analyzing them. This SOD fitting for the abbreviated isothermal chemical denaturation further fulfills data analysis methods on the insufficient data sets conducted in the two prevalent protein stability studies.

Colorectal cancer (CRC) is widespread with significant mortality. Both inherited and sporadic mutations in various signaling pathways influence the development and progression of the cancer. Identifying genetic mutations in CRC is important for optimal patient treatment and many approaches currently exist to uncover these mutations, including next-generation sequencing (NGS) and commercially available kits. In the present study, we used a semiconductor-based targeted DNA-sequencing approach to sequence and identify genetic mutations in 91 human rectal cancer samples. Analysis revealed frequent mutations in KRAS (58.2%), TP53 (28.6%), APC (16.5%), FBXW7 (9.9%) and PIK3CA (9.9%), and additional mutations in BRAF, CTNNB1, ERBB2 and SMAD4 were also detected at lesser frequencies. Thirty-eight samples (41.8%) also contained two or more mutations, with common combination mutations occurring between KRAS and TP53 (42.1%), and KRAS and APC (31.6%). DNA sequencing for individual cancers is of clinical importance for targeted drug therapy and the advantages of such targeted gene sequencing over other NGS platforms or commercially available kits in sensitivity, cost and time effectiveness may aid clinicians in treating CRC patients in the near future.

Breast cancer is the most common malignancy and the leading cause of cancer deaths in women worldwide. While specific genetic mutations have been linked to 5–10% of breast cancer cases, other environmental and epigenetic factors influence the development and progression of the cancer. Since unique mutations patterns have been observed in individual cancer samples, identification and characterization of the distinctive breast cancer molecular profile is needed to develop more effective target therapies. Until recently, identifying genetic cancer mutations via personalized DNA sequencing was impractical and expensive. The recent technological advancements in next-generation DNA sequencing, such as the semiconductor-based Ion Torrent sequencing platform, has made DNA sequencing cost and time effective with more reliable results. Using the Ion Torrent Ampliseq Cancer Panel, we sequenced 737 loci from 45 cancer-related genes to identify genetic mutations in 105 human breast cancer samples. The sequencing analysis revealed missense mutations in PIK3CA, and TP53 genes in the breast cancer samples of various histologic types. Thus, this study demonstrates the necessity of sequencing individual human cancers in order to develop personalized drugs or combination therapies to effectively target individual, breast cancer-specific mutations.

Lung cancer is the most common malignancy and the leading cause of cancer deaths worldwide. While smoking is by far the leading cause of lung cancer, other environmental and genetic factors influence the development and progression of the cancer. Since unique mutations patterns have been observed in individual cancer samples, identification and characterization of the distinctive lung cancer molecular profile is essential for developing more effective, tailored therapies. Until recently, personalized DNA sequencing to identify genetic mutations in cancer was impractical and expensive. The recent technological advancements in next-generation DNA sequencing, such as the semiconductor-based Ion Torrent sequencing platform, has made DNA sequencing cost and time effective with more reliable results. Using the Ion Torrent Ampliseq Cancer Panel, we sequenced 737 loci from 45 cancer-related genes to identify genetic mutations in 76 human lung cancer samples. The sequencing analysis revealed missense mutations in KRAS, EGFR, and TP53 genes in the breast cancer samples of various histologic types. Thus, this study demonstrates the necessity of sequencing individual human cancers in order to develop personalized drugs or combination therapies to effectively target individual, breast cancer-specific mutations.

Identifying gene mutations in individual tumors is critical to improve the efficacy of cancer therapy by matching targeted drugs to specific mutations. Gastrointestinal stromal tumors (GIST) are stromal or mesenchymal subepithelial neoplasms affecting the gastrointestinal tract and frequently contain activating gene mutations in either KIT or platelet-derived growth factor A (PDGFRA). Although GIST is highly responsive to several selective tyrosine kinase inhibitors, combined use of inhibitors targeting other mutations is needed to further prolong survival in patients with GIST. In this study, we aim to screen and identify genetic mutations in GIST for targeted therapy using the new Ion Torrent next-generation sequencing platform. Utilizing the Ion Ampliseq Cancer Panel, we sequenced 737 loci from 45 cancer-related genes using DNA extracted from formalin-fixed and paraffin-embedded (FFPE) samples of 121 human gastrointestinal stromal tumors, set up stringent parameters for reliable variant calling by filtering out potential raw base calling errors, and identified frequent mutations in the KIT gene. This study demonstrates the utility of using Ion Torrent sequencing to efficiently identify human cancer mutations. This may provide a molecular basis for clinically developing new drugs targeting these gene mutations for GIST therapy.

Gastric cancer is the one of the major causes of cancer-related death, especially in Asia. Gastric adenocarcinoma, the most common type of gastric cancer, is heterogeneous and its incidence and cause varies widely with geographical regions, gender, ethnicity, and diet. Since unique mutations have been observed in individual human cancer samples, identification and characterization of the molecular alterations underlying individual gastric adenocarcinomas is a critical step for developing more effective, personalized therapies. Until recently, identifying genetic mutations on an individual basis by DNA sequencing remained a daunting task. Recent advances in new next-generation DNA sequencing technologies, such as the semiconductor-based Ion Torrent sequencing platform, makes DNA sequencing cheaper, faster, and more reliable. In this study, we aim to identify genetic mutations in the genes which are targeted by drugs in clinical use or are under development in individual human gastric adenocarcinoma samples using Ion Torrent sequencing. We sequenced 737 loci from 45 cancer-related genes in 238 human gastric adenocarcinoma samples using the Ion Torrent Ampliseq Cancer Panel. The sequencing analysis revealed a high occurrence of mutations along the TP53 locus (9.7%) in our sample set. Thus, this study indicates the utility of a cost and time efficient tool such as Ion Torrent sequencing to screen cancer mutations for the development of personalized cancer therapy.

Lung cancer remains the most prevalent malignancy and the primary cause of cancer-related deaths worldwide. Unique mutations patterns can be found in lung cancer subtypes, in individual cancers, or within a single tumor, and drugs that target these genetic mutations and signal transduction pathways are often beneficial to patients. In this study, we used the Ion Torrent AmpliSeq Cancer Panel to sequence 737 loci from 45 cancer-related genes and oncogenes to identify genetic mutations in 48 formalin-fixed, paraffin-embedded (FFPE) human lung cancer samples from Chinese patients. We found frequent mutations in EGFR, KRAS, PIK3CA, and TP53 genes. Moreover, we observed that a portion of the lung cancer samples harbored two or more mutations in these key genes. This study demonstrates the feasibility of using the Ion Torrent sequencing to efficiently identify genetic mutations in individual tumors for targeted lung cancer therapy.

Identifying mutations in individual tumor is critical to improve the efficacy of cancer therapy by matching targeted drugs to specific mutations. Gastrointestinal stromal tumor (GIST) is a stromal or mesenchymal subepithelial neoplasm affecting the gastrointestinal tract and it frequently contains an activating mutation in either the KIT or platelet-derived growth factor A (PDGFRA) gene. Although GIST is highly responsive to several selective tyrosine kinase inhibitors, combined use of inhibitors targeting other mutations is needed to further prolong survival. In this study, we aim to screen and identify genetic mutations in GIST for targeted therapy using the new Ion Torrent next-generation sequencing platform. By utilizing the Ion Ampliseq Cancer Panel, we sequenced 737 loci from 45 cancer-related genes from DNA extracted from formalin-fixed and paraffin-embedded (FFPE) samples of 125 human gastrointestinal stromal tumors, set up stringent parameters for reliable variant calling by filtering out potential raw base calling errors, and identified frequent mutations in PDGFRA, KIT, APC, MLH1, MET, RET, and PIK3CA. Moreover, frequent missense mutations were detected in MLH1 (12%), MET (16.80%), KIT (33.60%), and PIK3CA (41.60%) genes. The mutations in MLH1 gene were frequently detected in the exons 2, while MET gene was frequently mutated in the exon 12. In addition, we identified missense mutations in other genes, including FLT3, KRAS, PDGFRA, and STK11, at lower frequencies. This study demonstrates the utility of using Ion Torrent sequencing to efficiently identify human cancer mutations. With the finding of the frequent mutations in MLH1 and MET genes, in addition to KIT, PDGFRA and PIK3CA, in GISTs, this study provides a molecular basis for clinically developing new drugs targeting these gene mutations for GIST therapy. Citation Format: Zhi Xu, Zhibin Hu, Xinying Huo, Chuanning Tang, Si-Yi Chen, Jinfei Chen. Molecular typing of Chinese gastrointestinal stromal tumors using a multigene next generation sequencing panel. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 1523. doi:10.1158/1538-7445.AM2014-1523

10544 Background: Identifying mutations in individual tumor is critical to improve the efficacy of cancer therapy by matching targeted drugs to specific mutations. Gastrointestinal stromal tumor (GIST) is a stromal or mesenchymal subepithelial neoplasm affecting the gastrointestinal tract and it frequently contains an activating mutation in either the KIT or platelet-derived growth factor A (PDGFRA) gene. Although GIST is highly responsive to several selective tyrosine kinase inhibitors, combined use of inhibitors targeting other mutations is needed to further prolong survival. Methods: In this study, we aim to screen and identify genetic mutations in GIST for targeted therapy using the new Ion Torrent next-generation sequencing platform. Results: By utilizing the Ion Ampliseq Cancer Panel, we sequenced 737 loci from 45 cancer-related genes from DNA extracted from formalin-fixed and paraffin-embedded (FFPE) samples of 125 human gastrointestinal stromal tumors, set up stringent parameters for reliable variant calling by filtering out potential raw base calling errors, and identified frequent mutations in PDGFRA, KIT, APC, MLH1, MET, RET, and PIK3CA. Moreover, frequent missense mutations were detected in MLH1 (12%), MET (16.8%), KIT (51.2%), and PIK3CA (41.6%) genes. The mutations in MLH1 gene were frequently detected in the exon 14, while MET gene was frequently mutated in the exon 2. In addition, we identified missense mutations in other genes, including FLT3, KRAS, PDGFRA, and STK11, at lower frequencies. Conclusions: This study demonstrates the utility of using Ion Torrent sequencing to efficiently identify human cancer mutations. With the finding of the frequent mutations in MLH1 and MET genes, in addition to KIT, PDGFRA and PIK3CA, in GISTs, this study provides a molecular basis for clinically developing new drugs targeting these gene mutations for GIST therapy.

Meta-predictors make predictions by organizing and processing the predictions produced by several other predictors in a defined problem domain. A proficient meta-predictor not only offers better predicting performance than the individual predictors from which it is constructed, but it also relieves experimentally researchers from making difficult judgments when faced with conflicting results made by multiple prediction programs. As increasing numbers of predicting programs are being developed in a large number of fields of life sciences, there is an urgent need for effective meta-prediction strategies to be investigated. We compiled four unbiased phosphorylation site datasets, each for one of the four major serine/threonine (S/T) protein kinase families-CDK, CK2, PKA and PKC. Using these datasets, we examined several meta-predicting strategies with 15 phosphorylation site predictors from six predicting programs: GPS, KinasePhos, NetPhosK, PPSP, PredPhospho and Scansite. Meta-predictors constructed with a generalized weighted voting meta-predicting strategy with parameters determined by restricted grid search possess the best performance, exceeding that of all individual predictors in predicting phosphorylation sites of all four kinase families. Our results demonstrate a useful decision-making tool for analysing the predictions of the various S/T phosphorylation site predictors. An implementation of these meta-predictors is available on the web at: http://MetaPred.umn.edu/MetaPredPS/.

Meta-prediction seeks to harness the combined strengths of multiple predicting programs with the hope of achieving predicting performance surpassing that of all existing predictors in a defined problem domain. We investigated meta-prediction for the four-compartment eukaryotic subcellular localization problem. We compiled an unbiased subcellular localization dataset of 1693 nuclear, cytoplasmic, mitochondrial and extracellular animal proteins from Swiss-Prot 50.2. Using this dataset, we assessed the predicting performance of 12 predictors from eight independent subcellular localization predicting programs: ELSPred, LOCtree, PLOC, Proteome Analyst, PSORT, PSORT II, SubLoc and WoLF PSORT. Gorodkin correlation coefficient (GCC) was one of the performance measures. Proteome Analyst is the best individual subcellular localization predictor tested in this four-compartment prediction problem, with GCC = 0.811. A reduced voting strategy eliminating six of the 12 predictors yields a meta-predictor (RAW-RAG-6) with GCC = 0.856, substantially better than all tested individual subcellular localization predictors (P = 8.2 x 10(-6), Fisher's Z-transformation test). The improvement in performance persists when the meta-predictor is tested with data not used in its development. This and similar voting strategies, when properly applied, are expected to produce meta-predictors with outstanding performance in other life sciences problem domains.