When building storage systems that aim to simultaneously provide robustness, scalability, and efficiency, one faces a fundamental tension, as higher robustness typically incurs higher costs and thus hurts both efficiency and scalability. My research shows that an approach to storage system design based on a simple principle—separating data from metadata—can yield systems that address elegantly and effectively that tension in a variety of settings. One observation motivates our approach: much of the cost paid by many strong protection techniques is incurred to detect errors. This observation suggests an opportunity: if we can build a low-cost oracle to detect errors and identify correct data, it may be possible to reduce the cost of protection without weakening its guarantees. This dissertation shows that metadata, if carefully designed, can serve as such an oracle and help a storage system protect its data with minimal cost. This dissertation shows how to effectively apply this idea in three very different systems: Gnothi—a storage replication protocol that combines the high availability of asynchronous replication and the low cost of synchronous replication for a small-scale block storage; Salus—a large-scale block storage with unprecedented guarantees in terms of consistency, availability, and durability in the face of a wide range of server failures; and Exalt—a tool to emulate a large storage system with 100 times fewer machines. / text

DNA replication is initiated at replication origins, which are temporally regulated within the S phase of the cell cycle so that some origins initiate early, while others initiate late. S. cerevisiae telomeres of normal length replicate late during the S phase. However, shortened telomeres replicate early – coupling the regulation of length to the replication timing. The mechanism through which telomere length regulates the activation time of nearby replication origins is unknown. In this thesis, I find that Tel1, a homolog of human ATM kinase, is required for early replication of short telomeres. In the absence of Tel1, short telomeres of the yku70Δ mutant no longer replicate early. I tested the role of Histone H2A(X) phosphorylation at Serine-129, as the target of Tel1 kinase, in regulation of telomeric replication times. However, preventing this phosphorylation had only a minor effect, so H2A(X) is unlikely to be the sole Tel1 target in telomeric replication regulation. On the other hand, deletion of Rif1 – a telomeric length regulator – showed complete epistasis to tel1Δ in telomeric replication regulation, implying that Rif1 acts downstream of Tel1 in the regulation of telomeric replication. Proteomic analysis revealed Tel1-dependent phosphorylation of Rif1 upon telomere shortening, implicating Rif1 as a direct downstream target of Tel1. However, mutation of the Tel1-phosphorylation sites on Rif1 had only a small effect on telomere replication times, indicating that phosphorylation of Rif1 by Tel1 is not the sole mechanism governing replication timing of short telomeres.

No description available.

Η αντιγραφή του DNA αποτελεί μια θεμελιώδη διαδικασία για κάθε μορφή ζωής. Στους ευκαρυωτικούς οργανισμούς, εξαιτίας του μεγάλου μεγέθους του γονιδιώματός τους, η αντιγραφή του DNA εκκινά από πολλαπλά σημεία προκειμένου να ολοκληρωθεί σε εύλογο χρονικό διάστημα. Οι περιοχές αυτές ονομάζονται αφετηρίες αντιγραφής και η μελέτη τους είναι σημαντική αφού είναι άρρηκτα συνδεδεμένες με την ακριβή ολοκλήρωση της αντιγραφής του DNA, διαδικασία ζωτικής σημασίας για το κύτταρο. Ενώ η λειτουργία των αφετηριών αντιγραφής είναι λίγο πολύ γνωστή, οι ακριβείς γονιδιωματικές δομές που συντελούν στο μηχανισμό επιλογής τους παραμένουν άγνωστες. Σκοπός αυτής της διπλωματικής εργασίας είναι η μελέτη, σε επίπεδο ολόκληρου του γονιδιώματος, των γονιδιωματικών περιοχών που αποτελούν αφετηρίες αντιγραφής του DNA και η εξαγωγή των χαρακτηριστικών που καθορίζουν την ιδιότητά τους να λειτουργούν ως αφετηρίες αντιγραφής του DNA. Ο ζυμομύκητας Schizosachharomyces pombe αποτελεί ιδανικό οργανισμό για τη μελέτη της διαδικασίας αντιγραφής του DNA, κυρίως εξαιτίας της ομοιότητας που διαθέτει με τους ανώτερους ευκαρυωτικούς οργανισμούς. Η παρούσα βιοπληροφορική ανάλυση εκτελεί επεξεργασία ολόκληρου του γονιδιώματος του S. pombe. Βασίζεται σε δύο πολύ πρόσφατα, αλλά διαφορετικών ερευνητικών ομάδων, πειράματα μικροσυστοιχειών στα οποία αναγνωρίστηκαν οι αφετηρίες αντιγραφής σε όλο το γονιδίωμα του ζυμομύκητα (Heichinger et al, 2006 και Hayashi et al, 2007). Συνδυάζοντας τα πειράματα αυτά, καταφέραμε να διαχωρίσουμε το σύνολο των διαγονιδιακών περιοχών του S. pombe σε κατηγορίες ανάλογα με την ικανότητα τους να εκκινούν την αντιγραφή του DNA. Έπειτα, ορίσαμε 3 νέα γονιδιωματικά παραμετρικά χαρακτηριστικά και μαζί με τη δημιουργία ενός συστήματος βελτιστοποίησης παραμέτρων, εξετάσαμε πώς πρέπει να διαμορφωθούν οι διάφορες παράμετροί τους ώστε να δημιουργούνται δομές που παρατηρούνται μονάχα στις αφετηρίες αντιγραφής. Για συγκεκριμένους συνδυασμούς παραμέτρων, η ταξινόμηση των διαγονιδιακών περιοχών με βάση τα καινούρια χαρακτηριστικά αγγίζει το 90% σε ευαισθησία, και το 77% σε θετικό προγνωστικό δείκτη. Επομένως, τα νέα γονιδιωματικά χαρακτηριστικά αφορούν δομές που παρατηρούνται σχεδόν εξολοκλήρου στις αφετηρίες αντιγραφής και σπάνια στις υπόλοιπες διαγονιδιακές περιοχές, παρέχοντας έτσι τον κύριο μηχανισμό επιλογής των αφετηριών αντιγραφής του DNA στο γονιδίωμα του S. pombe. / DNA replication constitutes an essential process for every life form. In eukaryotic organisms which are characterized by large genome size, DNA replication initiates from multiple points along the genome, so that it is completed within the allocated time. These genomic sites are called replication origins and their study is important as their selection and timely activation is pivotal for the maintenance of genomic integrity. Despite extensive studies from several laboratories, the features that specify an origin remain elusive, especially in higher eukaryotes. The purpose of this thesis is a genome-wide study of the genomic areas that function as replication origins and the extraction of the genomic features that determine this activity. Schizosachharomyces pombe (S. pombe, fission yeast) is an ideal organism for the study of the DNA replication procedure as it shares several common features with higher eukaryotic organisms. In this bioinformatics study, a genome-wide analysis of the S. pombe genome is performed. It is based on two very recent reports from different teams using microarray experiments, in which genome-wide identification of the S. pombe replication origins took place (Heichinger et al, 2006 and Hayashi et al, 2007). Combining these two experiments we managed to separate the fission yeast inter-genic regions in categories depending on their ability to function as replication origins. We then defined 3 new parametric genomic features, created a framework for solving the parameter estimation problem and analyzed how these parameters should be defined so that the specified structures are solely observed in genomic sites that function as replication origins. We observed that for certain parametric combinations, the classification of the intergenic regions as replication origins and as intergenic regions showing no origin activity reached 90% in sensitivity and 77% in positive predictive value. Therefore, the new genomic features identified through this study represent structures that are almost always and solely observed in intergenic regions showing replication activity, and are likely to provide the main genomic mechanism of origin selection in the fission yeast genome.

No description available.

The 1918 “Spanish Flu” pandemic was cause by an Influenza A virus that infected 500 million people and nearly 50 million people died. Influenza viruses utilize a RNA-Dependant RNA Polymerase (RdRp) complex, composed of the PA, PB1 and PB2 proteins along with the viral nucleoprotein (NP) to mediate viral transcription and replication. The 1918 PB1 gene has been linked to increased virulence in mice and ferrets. We have investigated the role of PB1 in the transcription and replication of the 1918 virus and its relation to pathogenicity by comparing its RdRp to the RdRp of a low virulence conventional H1N1 human virus isolate, A/Canada/RV733/2007 (RV733). Contrary to previous studies, our dual-luciferase reporter assay revealed the 1918 RdRp had lower transcriptional activity than the RV733 RdRp in vitro. The 1918 NP seems to be the key determinant for the difference in transcriptional activity of the 1918 and RV733 RdRp complexes. The 1918 PB1 in the RV733 RdRp maintained high reporter expression while the RV733 PB1 in the 1918 RdRp abolished reporter expression. 1918/RV733 chimeric PB1 proteins were also generated and evaluated with the reporter assay. Recombinant RV733/1918 viruses were generated by reverse genetics and we determined that PB1 is a key determinant of the high growth phenotype of the 1918 virus, but only a minor contributor to pathogenicity. A novel role for the 1918 NP in the high growth phenotype and pathogenicity of the 1918 virus is also described.

Hepatitis B virus genotype G (HBV/G) is a unique genotype of HBV which contains a 36-nucleotide insertion in the Core gene as well as 2 mutations that lead to stop codons in the Pre-Core coding region. Chronic infection with HBV/G is not known to occur without a co-infecting HBV genotype, suggesting that it is defective on its own. This study aims to look at the replication capacity of HBV/G, HBV/A, and HBV/A/G recombinant strains circulating in Canada and to determine the relationship between co-infecting strains. Four full-length HBV genomes were isolated from 2 different patients and transiently transfected into the HepG2 human hepatoma cell line for phenotypic analysis of each strain. HBV/G, HBV/A and HBV/A/G recombinant strains were isolated from Patient 1, while a different HBV/A/G recombinant strain was isolated from Patient 2. HBV replication capacity was measured using a quantitative real time PCR assay. Markers of replication, such as secreted HBsAg and HBeAg, intracellular core particles and replicative DNA intermediates were measured by ELISA, Western blot and Southern blot, respectively. HBV/G demonstrated a higher replicative capability, relative to its co-infecting strains, while both HBV/A/G strains had levels of secreted HBV DNA greater than HBV/A alone, suggesting a modulating effect due to recombination. Replication marker levels revealed possible reasons for a co-infection requirement during HBV/G infection such as HBeAg for chronicity. These observations demonstrate the potential interactions of HBV/G with its co-infecting HBV genotype and provide the first reported phenotypic analysis of a HBV recombinant.

No description available.

<p>Constitutive stable DNA replication (cSDR) is an alternative mode of replication initiation in Escherichia coli. cSDR is active in rnhA and recG mutants, which lack proteins that remove DNA-RNA hybrids called R-loops. The mechanism for cSDR initiation, therefore, is thought to involve these R-loop structures, which are proposed to form at specific locations known as oriK sites on the chromosome. Thus far, oriK sites have only been mapped to broad, 100-200 kb regions on the chromosome, so the specific elements involved in initiation are still unknown. My research focused on localizing the oriK sites on the chromosome, specifically those in the terminus region, where two of the major oriK sites had previously been mapped. We used two-dimensional gel electrophoresis (Friedman & Brewer, 1995) to analyze the replication forks that are blocked at the innermost Ter sites at the terminus, and found that elevated levels of replication forks are blocked at the Ter sites in rnhA mutants. We also used microarray and deep sequencing analysis to determine that there is a major location of oriK activity in the chromosome, located in the region between TerA and TerC. Furthermore, we also studied the use of the activation-induced deaminase (AID) enzyme as a tool for identifying regions of R-loop formation in the chromosome, and learned about its properties in the process.</p> / Dissertation

Lazy replication with snapshot isolation (SI) has emerged as a popular choice for distributed databases. However, lazy replication requires the execution of update transactions at one (master) site so that it is relatively easy for a total SI order to be determined for consistent installation of updates in the lazily replicated system. We propose a set of techniques that support update transaction execution over multiple partitioned sites, thereby allowing the master to scale. Our techniques determine a total SI order for update transactions over multiple master sites without requiring global coordination in the distributed system, and ensure that updates are installed in this order at all sites to provide consistent and scalable replication with SI. We have built our techniques into PostgreSQL and demonstrate their effectiveness through experimental evaluation.