Fady R. Mohareb

A comparison of artificial neural networks and partial least squares modelling for the rapid detection of the microbial spoilage of beef fillets based on Fourier transform infrared spectral fingerprints.

A series of partial least squares (PLS) models were employed to correlate spectral data from FTIR analysis with beef fillet spoilage during aerobic storage at different temperatures (0, 5, 10, 15, and 20 °C) using the dataset presented by Argyri et al. (2010). The performance of the PLS models was compared with a three-layer feed-forward artificial neural network (ANN) developed using the same dataset. FTIR spectra were collected from the surface of meat samples in parallel with microbiological analyses to enumerate total viable counts. Sensory evaluation was based on a three-point hedonic scale classifying meat samples as fresh, semi-fresh, and spoiled. The purpose of the modelling approach employed in this work was to classify beef samples in the respective quality class as well as to predict their total viable counts directly from FTIR spectra. The results obtained demonstrated that both approaches showed good performance in discriminating meat samples in one of the three predefined sensory classes. The PLS classification models showed performances ranging from 72.0 to 98.2% using the training dataset, and from 63.1 to 94.7% using independent testing dataset. The ANN classification model performed equally well in discriminating meat samples, with correct classification rates from 98.2 to 100% and 63.1 to 73.7% in the train and test sessions, respectively. PLS and ANN approaches were also applied to create models for the prediction of microbial counts. The performance of these was based on graphical plots and statistical indices (bias factor, accuracy factor, root mean square error). Furthermore, results demonstrated reasonably good correlation of total viable counts on meat surface with FTIR spectral data with PLS models presenting better performance indices compared to ANN.

MRMaid 2.0: Mining PRIDE for Evidence-Based SRM Transitions

Selected reaction monitoring (SRM) is becoming the tool of choice for targeted quantitative proteomics. The fundamental principle of proteomic SRM is that, for a given protein of interest, there is a set of peptides that are unique to that protein. The characteristic retention time (RT), and intact peptide m/z of these so-called proteotypic peptides are then programmed into the mass spectrometer, along with the m/z of high-intensity product ions for targeted quantitation. The particular combination of RT, peptide m/z, and product m/z for a given peptide is referred to as a transition. Selection of the most appropriate set of transitions for a given set of proteins is crucial to any SRM experiment. We previously developed the web-based MRMaid tool, which suggested the optimal transitions for a given human protein by mining spectral evidence from a small in-house database. In this article we present a completely new implementation of MRMaid, which offers substantial improvements over the original. The new version, MRMaid 2.0, uses spectra from the EBIs PRIDE database, which massively increases the coverage and quality of transitions. Transition lists can now be generated for multiple proteins simultaneously, edited within the web browser, and exported for laboratory use.

MRMaid: The SRM Assay Design Tool for Arabidopsis and Other Species

Selected reaction monitoring (SRM), sometimes called multiple reaction monitoring (MRM), is becoming the tool of choice for targeted quantitative proteomics in the plant science community. Key to a successful SRM experiment is prior identification of the distinct peptides for the proteins of interest and the determination of the so-called transitions that can be programmed into an LC-MS/MS to monitor those peptides. The transition for a given peptide comprises the intact peptide m/z and a high intensity product ion that can be monitored at a characteristic retention time (RT). To aid the design of SRM experiments, several online tools and databases have been produced to help researchers select transitions for their proteins of interest, but many of these tools are limited to the most popular model organisms such as human, yeast, and mouse or require the experimental acquisition of local spectral libraries. In this paper we present MRMaid1, a web-based SRM assay design tool whose transitions are generated by mining the millions of identified peptide spectra held in the EBI’s PRIDE database. By using data from this large public repository, MRMaid is able to cover a wide range of species that can increase as the coverage of PRIDE grows. In this paper MRMaid transitions for 25 Arabidopsis thaliana proteins are evaluated experimentally, and found capable of quantifying 23 of these proteins. This performance was found to be comparable with the more time consuming approach of designing transitions using locally acquired orbitrap data, indicating that MRMaid is a valuable tool for targeted Arabidopsis proteomics.

PhyTB: Phylogenetic tree visualisation and sample positioning for M. tuberculosis

BackgroundPhylogenetic-based classification of M. tuberculosis and other bacterial genomes is a core analysis for studying evolutionary hypotheses, disease outbreaks and transmission events. Whole genome sequencing is providing new insights into the genomic variation underlying intra- and inter-strain diversity, thereby assisting with the classification and molecular barcoding of the bacteria. One roadblock to strain investigation is the lack of user-interactive solutions to interrogate and visualise variation within a phylogenetic tree setting.ResultsWe have developed a web-based tool called PhyTB (http://pathogenseq.lshtm.ac.uk/phytblive/index.php) to assist phylogenetic tree visualisation and identification of M. tuberculosis clade-informative polymorphism. Variant Call Format files can be uploaded to determine a sample position within the tree. A map view summarises the geographical distribution of alleles and strain-types. The utility of the PhyTB is demonstrated on sequence data from 1,601 M. tuberculosis isolates.ConclusionPhyTB contextualises M. tuberculosis genomic variation within epidemiological, geographical and phylogenic settings. Further tool utility is possible by incorporating large variants and phenotypic data (e.g. drug-resistance profiles), and an assessment of genotype-phenotype associations. Source code is available to develop similar websites for other organisms (http://sourceforge.net/projects/phylotrack).

Study of microRNAs-21/221 as potential breast cancer biomarkers in Egyptian women

microRNAs (miRNAs) play an important role in cancer prognosis. They are small molecules, approximately 17-25 nucleotides in length, and their high stability in human serum supports their use as novel diagnostic biomarkers of cancer and other pathological conditions. In this study, we analyzed the expression patterns of miR-21 and miR-221 in the serum from a total of 100 Egyptian female subjects with breast cancer, fibroadenoma, and healthy control subjects. Using microarray-based expression profiling followed by real-time polymerase chain reaction validation, we compared the levels of the two circulating miRNAs in the serum of patients with breast cancer (n=50), fibroadenoma (n=25), and healthy controls (n=25). The miRNA SNORD68 was chosen as the housekeeping endogenous control. We found that the serum levels of miR-21 and miR-221 were significantly overexpressed in breast cancer patients compared to normal controls and fibroadenoma patients. Receiver Operating Characteristic (ROC) curve analysis revealed that miR-21 has greater potential in discriminating between breast cancer patients and the control group, while miR-221 has greater potential in discriminating between breast cancer and fibroadenoma patients. Classification models using k-Nearest Neighbor (kNN), Naïve Bayes (NB), and Random Forests (RF) were developed using expression levels of both miR-21 and miR-221. Best classification performance was achieved by NB Classification models, reaching 91% of correct classification. Furthermore, relative miR-221 expression was associated with histological tumor grades. Therefore, it may be concluded that both miR-21 and miR-221 can be used to differentiate between breast cancer patients and healthy controls, but that the diagnostic accuracy of serum miR-21 is superior to miR-221 for breast cancer prediction. miR-221 has more diagnostic power in discriminating between breast cancer and fibroadenoma patients. The overexpression of miR-221 has been associated with the breast cancer grade. We also demonstrated that the combined expression of miR-21 and miR-221can be successfully applied as breast cancer biomarkers.

Resequencing at ≥40-Fold Depth of the Parental Genomes of a Solanum lycopersicum × S. pimpinellifolium Recombinant Inbred Line Population and Characterization of Frame-Shift InDels That Are Highly Likely to Perturb Protein Function.

A recombinant in-bred line population derived from a cross between Solanum lycopersicum var. cerasiforme (E9) and S. pimpinellifolium (L5) has been used extensively to discover quantitative trait loci (QTL), including those that act via rootstock genotype, however, high-resolution single-nucleotide polymorphism genotyping data for this population are not yet publically available. Next-generation resequencing of parental lines allows the vast majority of polymorphisms to be characterized and used to progress from QTL to causative gene. We sequenced E9 and L5 genomes to 40- and 44-fold depth, respectively, and reads were mapped to the reference Heinz 1706 genome. In L5 there were three clear regions on chromosome 1, chromosome 4, and chromosome 8 with increased rates of polymorphism. Two other regions were highly polymorphic when we compared Heinz 1706 with both E9 and L5 on chromosome 1 and chromosome 10, suggesting that the reference sequence contains a divergent introgression in these locations. We also identified a region on chromosome 4 consistent with an introgression from S. pimpinellifolium into Heinz 1706. A large dataset of polymorphisms for the use in fine-mapping QTL in a specific tomato recombinant in-bred line population was created, including a high density of InDels validated as simple size-based polymerase chain reaction markers. By careful filtering and interpreting the SnpEff prediction tool, we have created a list of genes that are predicted to have highly perturbed protein functions in the E9 and L5 parental lines.

Biochemical Profile of Heritage and Modern Apple Cultivars and Application of Machine Learning Methods To Predict Usage, Age, and Harvest Season

The present study represents the first major attempt to characterize the biochemical profile in different tissues of a large selection of apple cultivars sourced from the United Kingdoms National Fruit Collection comprising dessert, ornamental, cider, and culinary apples. Furthermore, advanced machine learning methods were applied with the objective to identify whether the phenolic and sugar composition of an apple cultivar could be used as a biomarker fingerprint to differentiate between heritage and mainstream commercial cultivars as well as govern the separation among primary usage groups and harvest season. A prediction accuracy of >90% was achieved with the random forest method for all three models. The results highlighted the extraordinary phytochemical potency and unique profile of some heritage, cider, and ornamental apple cultivars, especially in comparison to more mainstream apple cultivars. Therefore, these findings could guide future cultivar selection on the basis of health-promoting phytochemical content.

Genomic variation in Plasmodium vivax malaria reveals regions under selective pressure.

Background Although Plasmodium vivax contributes to almost half of all malaria cases outside Africa, it has been relatively neglected compared to the more deadly P. falciparum. It is known that P. vivax populations possess high genetic diversity, differing geographically potentially due to different vector species, host genetics and environmental factors. Results We analysed the high-quality genomic data for 46 P. vivax isolates spanning 10 countries across 4 continents. Using population genetic methods we identified hotspots of selection pressure, including the previously reported MRP1 and DHPS genes, both putative drug resistance loci. Extra copies and deletions in the promoter region of another drug resistance candidate, MDR1 gene, and duplications in the Duffy binding protein gene (PvDBP) potentially involved in erythrocyte invasion, were also identified. For surveillance applications, continental-informative markers were found in putative drug resistance loci, and we show that organellar polymorphisms could classify P. vivax populations across continents and differentiate between Plasmodia spp. Conclusions This study has shown that genomic diversity that lies within and between P. vivax populations can be used to elucidate potential drug resistance and invasion mechanisms, as well as facilitate the molecular barcoding of the parasite for surveillance applications.

Global genetic diversity of var2csa in Plasmodium falciparum with implications for malaria in pregnancy and vaccine development

Malaria infection during pregnancy, caused by the sequestering of Plasmodium falciparum parasites in the placenta, leads to high infant mortality and maternal morbidity. The parasite-placenta adherence mechanism is mediated by the VAR2CSA protein, a target for natural occurring immunity. Currently, vaccine development is based on its ID1-DBL2Xb domain however little is known about the global genetic diversity of the encoding var2csa gene, which could influence vaccine efficacy. In a comprehensive analysis of the var2csa gene in >2,000 P. falciparum field isolates across 23 countries, we found that var2csa is duplicated in high prevalence (>25%), African and Oceanian populations harbour a much higher diversity than other regions, and that insertions/deletions are abundant leading to an underestimation of the diversity of the locus. Further, ID1-DBL2Xb haplotypes associated with adverse birth outcomes are present globally, and African-specific haplotypes exist, which should be incorporated into vaccine design.

Visceral fat mass as a novel risk factor for predicting gestational diabetes in obese pregnant women

Objective To develop a model to predict gestational diabetes mellitus incorporating classical and a novel risk factor, visceral fat mass. Methods Three hundred two obese non-diabetic pregnant women underwent body composition analysis at booking by bioimpedance analysis. Of this cohort, 72 (24%) developed gestational diabetes mellitus. Principal component analysis was initially performed to identify possible clustering of the gestational diabetes mellitus and non-GDM groups. A machine learning algorithm was then applied to develop a GDM predictive model utilising random forest and decision tree modelling. Results The predictive model was trained on 227 samples and validated using an independent testing subset of 75 samples where the model achieved a validation prediction accuracy of 77.53%. According to the decision tree developed, visceral fat mass emerged as the most important variable in determining the risk of gestational diabetes mellitus. Conclusions We present a model incorporating visceral fat mass, which is a novel risk factor in predicting gestational diabetes mellitus in obese pregnant women.