Design and implementation of bioinformatics tools for large scale genome annotation

The continuous increase of genome sequencing projects produced a huge amount of data in the last 10 years: currently more than 600 prokaryotic and 80 eukaryotic genomes are fully sequenced and publically available. However the sole sequencing process of a genome is able to determine just raw nucleotide sequences. This is only the first step of the genome annotation process that will deal with the issue of assigning biological information to each sequence. The annotation process is done at each different level of the biological information processing mechanism, from DNA to protein, and cannot be accomplished only by in vitro analysis procedures resulting extremely expensive and time consuming when applied at a this large scale level. Thus, in silico methods need to be used to accomplish the task. The aim of this work was the implementation of predictive computational methods to allow a fast, reliable, and automated annotation of genomes and proteins starting from aminoacidic sequences. The first part of the work was focused on the implementation of a new machine learning based method for the prediction of the subcellular localization of soluble eukaryotic proteins. The method is called BaCelLo, and was developed in 2006. The main peculiarity of the method is to be independent from biases present in the training dataset, which causes the over‐prediction of the most represented examples in all the other available predictors developed so far. This important result was achieved by a modification, made by myself, to the standard Support Vector Machine (SVM) algorithm with the creation of the so called Balanced SVM. BaCelLo is able to predict the most important subcellular localizations in eukaryotic cells and three, kingdom‐specific, predictors were implemented. In two extensive comparisons, carried out in 2006 and 2008, BaCelLo reported to outperform all the currently available state‐of‐the‐art methods for this prediction task. BaCelLo was subsequently used to completely annotate 5 eukaryotic genomes, by integrating it in a pipeline of predictors developed at the Bologna Biocomputing group by Dr. Pier Luigi Martelli and Dr. Piero Fariselli. An online database, called eSLDB, was developed by integrating, for each aminoacidic sequence extracted from the genome, the predicted subcellular localization merged with experimental and similarity‐based annotations. In the second part of the work a new, machine learning based, method was implemented for the prediction of GPI‐anchored proteins. Basically the method is able to efficiently predict from the raw aminoacidic sequence both the presence of the GPI‐anchor (by means of an SVM), and the position in the sequence of the post‐translational modification event, the so called ω‐site (by means of an Hidden Markov Model (HMM)). The method is called GPIPE and reported to greatly enhance the prediction performances of GPI‐anchored proteins over all the previously developed methods. GPIPE was able to predict up to 88% of the experimentally annotated GPI‐anchored proteins by maintaining a rate of false positive prediction as low as 0.1%. GPIPE was used to completely annotate 81 eukaryotic genomes, and more than 15000 putative GPI‐anchored proteins were predicted, 561 of which are found in H. sapiens. In average 1% of a proteome is predicted as GPI‐anchored. A statistical analysis was performed onto the composition of the regions surrounding the ω‐site that allowed the definition of specific aminoacidic abundances in the different considered regions. Furthermore the hypothesis that compositional biases are present among the four major eukaryotic kingdoms, proposed in literature, was tested and rejected. All the developed predictors and databases are freely available at: BaCelLo http://gpcr.biocomp.unibo.it/bacello eSLDB http://gpcr.biocomp.unibo.it/esldb GPIPE http://gpcr.biocomp.unibo.it/gpipe

Imaging di risonanza magnetica nucleare nello studio del tumore della prostata

Aim of the present study was to evaluate the accuracy of transrectal ultrasound biopsy (TRUS-biopsy) directed to regions with abnormal MRI and/or MRSI (magnetic resonance spectroscopic imaging ) for both the transition (TZ) and the peripheral (PZ) zones in patients who presented with persistent suspect for prostate cancer and with prior negative biopsy. We also evaluated relationship between MRSI results and histopathological findings of biopsy. 54 patients with the aforementioned characteristics underwent MRI/MRSI at least 6 months after prior negative biopsy; interval between MRI/3D-MRSI and the further TRUS-biopsy was less than 3 months. The prostate was divided in 12 regions both for imaging interpretation and biopsy. Moreover one to three cores more were taken from each region with abnormal MRI and/or 3D-MRSI. Twenty-two out of 54 patients presented cancer at MRI/MRSI-directed-TRUS-biopsy. On a patient basis the highest accuracy was obtained by assigning malignancy on a positive finding with MRSI and MRI even though it was not significantly greater than that obtained using MRI alone (area under the ROC curve, AUC: 0.723 vs. 0.676). On a region (n=648) basis the best accuracy was also obtained by considering positive both MRSI and MRI for PZ (0.768) and TZ (0.822). Twenty-eight per cent of cores with prostatitis were false positive findings on MRSI, whereas only 2.7% of benign prostatic hyperplasia was false positive. In conclusion the accuracy of MRI/MRSI-directed biopsies in localization of prostate cancer is good in patient and region analyses. The combination of both MRI and MRSI results makes TRUS-biopsy more accurate particularly in the TZ (0.822) for patients with prior negative biopsies. Histopathological analysis showed that the main limitation of MRSI is the percentage of false positive findings due to prostatitis.

Identificazione e dosaggio di marcatori molecolari dell'endometriosi nel sangue periferico

Purpose: to quantify the mRNA levels of MMP-3, MMP-9, VEGF and Survivin in peripheral blood and the serum levels of CA-125, Ca19-9 in women with and without endometriosis and to investigate the performance of these markers to differentiate between deep and ovarian endometriosis. Methods: a case controls study enrolled a series of 60 patients. Twenty controls have been matched with 20 cases of ovarian and 20 cases of deep endometriosis. Univariable and multivariable performance of serum CA125 and CA19-9, mRNA for Survivin, MMP9, MMP3 and VEGF genes have been evaluated by means of ROC curves and logistic regression respectively. Results: No difference in markers concentration were detected between ovarian and deep endometriosis. In comparison with controls serum CA19 and CA125 yielded the better sensitivity followed by mRNA for Survivin gene (81.5%, 51.9% and 7.5% at 10% false positive rate respectively). Multivariable estimated odds of endometriosis yielded a sensitivity of 87% at the same false positive rate. Conclusions: A combination of serum and molecular markers could allow a better diagnosis of endometriosis.

Characterization of the Caspr2 and NLGN2 ligands: a proteomic and biochemical approach

Autism Spectrum Disorder (ASD) is a range of early-onset conditions classified as neurodevelopmental disorders, characterized by deficits in social interactions and communication, as well as by restricted interest and repetitive behaviors. Among the proteins associated with this spectrum of disease there are Caspr2, α-NRXN1, NLGN1-4. Caspr2 is involved in the clustering of K+ channels at the juxtaparanodes, where it is proposed to bind TAG-1. Recent works reported a synaptic localization of Caspr2, but little is know on its role in this compartment. NRXNs and their ligand NLGNs, instead, have a well-defined role in the formation and maintenance of synapses. Among the neuroligins, NLGN2 binds NRXNs with the lowest affinity, suggesting that it could have other not yet characterized ligands. The aim of this work was to better characterize the binding of Caspr2 to TAG-1 and to identify new potential binding partner for Caspr2 and NLGN2. Unexpectedly, using Isothermal Titration Calorimetry and co-immunoprecipitation experiments the direct association of the first two proteins could not be verified and the results indicate that the first evidences reporting it were biased by false-positive artifacts. These findings, together with the uncharacterized synaptic localization of Caspr2, made the identification of new potential binding partners for this protein necessary. To find new proteins that associate with Caspr2 and NLGN2, affinity chromatography in tandem with mass spectrometry experiments were performed. Interestingly, about 25 new potential partners were found for these two proteins and NLGN1, that was originally included as a control: 5 of those, namely SFRP1, CLU, APOE, CNTN1 and TNR, were selected for further investigations. Only the association of CLU to NLGN2 was confirmed. In the future, screenings of the remaining candidates have to be carried out and the functional role for the proposed NLGN2-CLU complex has to be studied.