A profile-based tool for finding pipeline parallelism in sequential programs

Traditional static analysis fails to auto-parallelize programs with a complex control and data flow. Furthermore, thread-level parallelism in such programs is often restricted to pipeline parallelism, which can be hard to discover by a programmer. In this paper we propose a tool that, based on profiling information, helps the programmer to discover parallelism. The programmer hand-picks the code transformations from among the proposed candidates which are then applied by automatic code transformation techniques.

This paper contributes to the literature by presenting a profiling tool for discovering thread-level parallelism. We track dependencies at the whole-data structure level rather than at the element level or byte level in order to limit the profiling overhead. We perform a thorough analysis of the needs and costs of this technique. Furthermore, we present and validate the belief that programs with complex control and data flow contain significant amounts of exploitable coarse-grain pipeline parallelism in the program’s outer loops. This observation validates our approach to whole-data structure dependencies. As state-of-the-art compilers focus on loops iterating over data structure members, this observation also explains why our approach finds coarse-grain pipeline parallelism in cases that have remained out of reach for state-of-the-art compilers. In cases where traditional compilation techniques do find parallelism, our approach allows to discover higher degrees of parallelism, allowing a 40% speedup over traditional compilation techniques. Moreover, we demonstrate real speedups on multiple hardware platforms.

On generating set index functions for randomized caches

Caches hide the growing latency of accesses to the main memory from the processor by storing the most recently used data on-chip. To limit the search time through the caches, they are organized in a direct mapped or set-associative way. Such an organization introduces many conflict misses that hamper performance. This paper studies randomizing set index functions, a technique to place the data in the cache in such a way that conflict misses are avoided. The performance of such a randomized cache strongly depends on the randomization function. This paper discusses a methodology to generate randomization functions that perform well over a broad range of benchmarks. The methodology uses profiling information to predict the conflict miss rate of randomization functions. Then, using this information, a search algorithm finds the best randomization function. Due to implementation issues, it is preferable to use a randomization function that is extremely simple and can be evaluated in little time. For these reasons, we use randomization functions where each randomized address bit is computed as the XOR of a subset of the original address bits. These functions are chosen such that they operate on as few address bits as possible and have few inputs to each XOR. This paper shows that to index a 2(m)-set cache, it suffices to randomize m+2 or m+3 address bits and to limit the number of inputs to each XOR to 2 bits to obtain the full potential of randomization. Furthermore, it is shown that the randomization function that we generate for one set of benchmarks also works well for an entirely different set of benchmarks. Using the described methodology, it is possible to reduce the implementation cost of randomization functions with only an insignificant loss in conflict reduction.

Designing computer architecture research workloads

MinneSPEC proposes reduced input sets that microprocessor designers can use to model representative short-running workloads. A four-step methodology verifies the program behavior similarity of these input sets to reference sets.

Circulating and synovial antibody profiling of juvenile arthritis patients by nucleic acid programmable protein arrays.

Introduction Juvenile idiopathic arthritis (JIA) is a heterogeneous disease characterized by chronic joint inflammation of unknown cause in children. JIA is an autoimmune disease and small numbers of auto-antibodies have been reported in JIA patients. The identification of antibody markers could improve the existing clinical management of patients. Methods A pilot study was performed on the application of a high-throughput platform, nucleic acid programmable protein arrays (NAPPA), to assess the levels of antibodies present in the systemic circulation and synovial joint of a small cohort of juvenile arthritis patients. Plasma and synovial fluid from ten JIA patients was screened for antibodies against 768 proteins on NAPPA. Results Quantitative reproducibility of NAPPA was demonstrated with >0.95 intra- and inter- array correlations. A strong correlation was also observed for the levels of antibodies between plasma and synovial fluid across the study cohort (r=0.96). Differences in the levels of 18 antibodies were revealed between sample types across all patients. Patients were segregated into two clinical subtypes with distinct antibody signatures by unsupervised hierarchical cluster analysis. Conclusions NAPPA provides a high-throughput quantitatively reproducible platform to screen for disease specific autoantibodies at the proteome level on a microscope slide. The strong correlation between the circulating antibody levels and those of the inflamed joint represents a novel finding and provides confidence to use plasma for discovery of autoantibodies in JIA, thus circumventing the challenges associated with joint aspiration. We expect that autoantibody profiling of JIA patients on NAPPA could yield antibody markers that can act as criteria to stratify patients, predict outcomes and understand disease etiology at the molecular level.

Use of NMR metabolomic plasma profiling methodologies to identify illicit growth-promoting administrations

Detection of growth-promoter use in animal production systems still proves to be an analytical challenge despite years of activity in the field. This study reports on the capability of NMR metabolomic profiling techniques to discriminate between plasma samples obtained from cattle treated with different groups of growth-promoting hormones (dexamethasone, prednisolone, oestradiol) based on recorded metabolite profiles. Two methods of NMR analysis were investigated—a Carr–Purcell–Meiboom–Gill (CPMG)-pulse sequence technique and a conventional 1H NMR method using pre-extracted plasma. Using the CPMG method, 17 distinct metabolites could be identified from the spectra. 1H NMR analysis of extracted plasma facilitated identification of 23 metabolites—six more than the alternative method and all within the aromatic region. Multivariate statistical analysis of acquired data from both forms of NMR analysis separated the plasma metabolite profiles into distinct sample cluster sets representative of the different animal study groups. Samples from both sets of corticosteroid-treated animals—dexamethasone and prednisolone—were found to be clustered relatively closely and had similar alterations to identified metabolite panels. Distinctive metabolite profiles, different from those observed within plasma from corticosteroid-treated animal plasma, were observed in oestradiol-treated animals and samples from these animals formed a cluster spatially isolated from control animal plasma samples. These findings suggest the potential use of NMR methodologies of plasma metabolite analysis as a high-throughput screening technique to aid detection of growth promoter use.

Validated ligand binding sites in CCK receptors. Next step: Computer aided design of novel CCK ligands

Computer-aided drug design becomes an important part of G-protein coupled receptors (GPCR) drug discovery process that is applied for improving the efficiency of derivation and optimization of novel ligands. It represents the combination of methods that-use-structural information of a receptor binding site of known ligands to design new ligands. In this report, we give a brief description of ligand binding sites in cholecystokinin and gastrin receptors (CK1R and CCK2R) which were delineated using experimental and computational methods, and then, we show how the validated ligand binding sites can be used to design and improve novel ligands. The translation of the knowledge of ligand-binding sites of different GPCRs to computer-aided design of novel ligands is summarized.

Genome-wide profiling of methylation identifies novel targets with aberrant hyper-methylation and reduced expression in low-risk myelodysplastic syndromes

Gene expression profiling signatures may be used to classify the subtypes of Myelodysplastic syndrome (MDS) patients. However, there are few reports on the global methylation status in MDS. The integration of genome-wide epigenetic regulatory marks with gene expression levels would provide additional information regarding the biological differences between MDS and healthy controls. Gene expression and methylation status were measured using high-density microarrays. A total of 552 differentially methylated CpG loci were identified as being present in low-risk MDS; hypermethylated genes were more frequent than hypomethylated genes. In addition, mRNA expression profiling identified 1005 genes that significantly differed between low-risk MDS and the control group. Integrative analysis of the epigenetic and expression profiles revealed that 66.7% of the hypermethylated genes were underexpressed in low-risk MDS cases. Gene network analysis revealed molecular mechanisms associated with the low-risk MDS group, including altered apoptosis pathways. The two key apoptotic genes BCL2 and ETS1 were identified as silenced genes. In addition, the immune response and micro RNA biogenesis were affected by the hypermethylation and underexpression of IL27RA and DICER1. Our integrative analysis revealed that aberrant epigenetic regulation is a hallmark of low-risk MDS patients and could have a central role in these diseases.

Pharmacogenomic Profiling and Pathway Analyses Identify MAPK-Dependent Migration as an Acute Response to SN38 in p53 Null and p53-Mutant Colorectal Cancer Cells.

The topoisomerase I inhibitor irinotecan is used to treat advanced colorectal cancer and has been shown to have p53-independent anticancer activity. The aim of this study was to identify the p53-independent signaling mechanisms activated by irinotecan. Transcriptional profiling of isogenic HCT116 p53 wild-type and p53 null cells was carried out following treatment with the active metabolite of irinotecan, SN38. Unsupervised analysis methods showed that p53 status had a highly significant impact on gene expression changes in response to SN38. Pathway analysis indicated that pathways involved in cell motility [adherens junction, focal adhesion, mitogen-activated protein kinase (MAPK), and regulation of the actin cytoskeleton] were significantly activated in p53 null cells, but not p53 wild-type cells, following SN38 treatment. In functional assays, SN38 treatment increased the migratory potential of p53 null and p53-mutant colorectal cancer cell lines, but not p53 wild-type lines. Moreover, p53 null SN38-resistant cells were found to migrate at a faster rate than parental drug-sensitive p53 null cells, whereas p53 wild-type SN38-resistant cells failed to migrate. Notably, cotreatment with inhibitors of the MAPK pathway inhibited the increased migration observed following SN38 treatment in p53 null and p53-mutant cells. Thus, in the absence of wild-type p53, SN38 promotes migration of colorectal cancer cells, and inhibiting MAPK blocks this potentially prometastatic adaptive response to this anticancer drug.

Pharmacological profiling of store-operated Ca(2+) entry in retinal arteriolar smooth muscle

Objective: Pharmacological profiling of store-operated Ca(2+) entry (SOCE) and molecular profiling of ORAI and TRPC expression in arterioles. 
Methods: Fura-2 based microfluorimetry was used to assess CPA-induced SOCE in rat retinal arteriolar myocytes. Arteriolar ORAI and TRP transcript expression were screened using RT-PCR. 
Results: SKF96365 and LOE908 blocked SOCE (IC(50) s of 1.2µM and 1.4µM, respectively). Gd(3+) and La(3+) potently inhibited SOCE (IC(50) s of 21nM and 42nM, respectively), but Ni(2+) showed lower potency (IC(50) = 11.6µM). 2-aminoethyldiphenyl borate (2APB) inhibited SOCE (IC(50) = 3.7µM) but enhanced basal influx (>100µM). Verapamil and nifedipine had no effect at concentrations that inhibit L-type Ca(2+) channels, but diltiazem inhibited SOCE by approximately 40% (=0.1µM). RT-PCR demonstrated transcript expression for ORAI 1, 2 and 3, and TRPC1, 3, 4 and 7. Transcripts for TRPV1 and 2, which are activated by 2APB, were also expressed. 
Conclusion: The pharmacological profile of SOCE in retinal arteriolar smooth muscle appears unique when compared to other vascular tissues. This suggests that the molecular mechanisms underlying SOCE can differ, even in closely related tissues. Taken together, the pharmacological and molecular data are most consistent with involvement of TRPC1 in SOCE, although involvement of ORAI or other TRPC channels cannot be excluded. © 2012 John Wiley & Sons Ltd.

Premiere of "Spaced Images with Noise and Lines" Performance at the Joinery, Dublin, at at Spatial Music Collective concert. Authors: Lyon, E. Keywords: (Spatial Audio, Electroacoustic Music, Computer Music, Algorithmic Music) Publication date: 07 Oct 2011

Performance at the Joinery, Dublin, at at Spatial Music Collective concert

General practitioner records on computer--handle with care

This 1 year prospective study involved nine general practitioners in an urban health centre who routinely record all patient contacts on computer. The study determines by comparison with a manual record how accurately doctors record laboratory investigations on computer and compares the effectiveness of three interventions in improving the completeness of computerized recording of presenting symptoms, problems/diagnoses and laboratory investigations. Recording was analysed for 1 month prior to and for two 1 month periods following each intervention. A control group was used. A total of 7983 patient contacts were analysed. Intervention led to an improvement in the recording of presenting symptoms and problems/diagnoses. Recording of investigations on the computer showed no improvement, remaining at one-third of the total in the treatment room book for both study and control doctors. The effectiveness of the different forms of intervention depended on both the aspect of the consultation considered and the familiarity of individual doctors with the method of data collection. Aspects considered less important required greater intervention to bring about a marked improvement, as did doctors relatively new to the practice. It may not be possible to get all aspects of the consultation recorded with the same degree of accuracy. This has implications for the accuracy of retrospective studies dependent on existing computerized data.

"The End" for piano and computer. Performed by Shiau-uen Ding. Composer's Voice Concert Series: 15 Minutes of Fame. Jan Hus Church, NYC. General information State: Published Organisations: School of Creative Arts, Sonic Arts, UOA35 - Music, Drama, Dance and Performing Arts Authors: Lyon, E.

Performed by Shiau-uen Ding. Composer's Voice Concert Series: 15 Minutes of Fame. Jan Hus Church, NYC.