Boston College

About: Boston College is a education organization based out in Boston, Massachusetts, United States. It is known for research contribution in the topics: Population & Poison control. The organization has 9749 authors who have published 25406 publications receiving 1105145 citations. The organization is also known as: BC.

Editorial Overview---The Role of Information Systems in Healthcare: Current Research and Future Trends

Abstract: Information systems have great potential to reduce healthcare costs and improve outcomes. The purpose of this special issue is to offer a forum for theory-driven research that explores the role of IS in the delivery of healthcare in its diverse organizational and regulatory settings. We identify six theoretically distinctive elements of the healthcare context and discuss how these elements increase the motivation for, and the salience of, the research results reported in the nine papers comprising this special issue. We also provide recommendations for future IS research focusing on the implications of technology-driven advances in three areas: social media, evidence-based medicine, and personalized medicine.

DiANNA: a web server for disulfide connectivity prediction

Abstract: Correctly predicting the disulfide bond topology in aproteinisofcrucialimportancefortheunderstandingofproteinfunctionandcanbeofgreathelpfortertiarypredictionmethods.Thewebserverhttp://clavius.bc.edu/~clotelab/DiANNA/ outputs the disulfide con-nectivitypredictiongiveninputofaproteinsequence.Thefollowingprocedureisperformed.First,PSIPREDis run to predict the protein’s secondary structure,then PSIBLAST is run against the non-redundantSwissProt to obtain a multiple alignment of the inputsequence.Thepredictedsecondarystructureandtheprofile arising from this alignment are used in thetraining phase of our neural network. Next, cysteineoxidation state is predicted, then each pair ofcysteines in the protein sequence is assigned a like-lihoodofformingadisulfidebond—thisisperformedby means of a novel architecture (diresidue neuralnetwork). Finally, Rothberg’s implementation ofGabow’s maximum weighted matching algorithm isapplied to diresidue neural network scores in ordertoproducethefinalconnectivityprediction.Ournovelneural network-based approach achieves resultsthat are comparable and in some cases better thanthe current state-of-the-art methods.INTRODUCTIONDisulfide bonds are covalently bonded sulfur atoms fromnonadjacent cysteine residues, which stabilize the proteinstructure and are often found in extracytoplasmatic proteins.The knowledge of cysteine connectivity (i.e. which, if any,pairsofcysteinesformabondinagivenproteinsequence)canreduce greatly the conformational space for protein structureprediction algorithms. Moreover, as shown by Chuang andco-workers (1), a similar disulfide connectivity pattern fre-quently implies astructural similarity even when the sequencesimilarity is undetectable. Notwithstanding, only a fewattempts have been made to solve this problem. In contrast,many methods have been developed for the related, butsimpler problem of cysteine oxidation state prediction, i.e.to determine the cysteines that are involved in a disulfidebond, without predicting the connectivity pattern. Recentmethods based on machine learning techniques have reachedan outstanding accuracy of 90% on certain test data (2–5).Inspiteofthis,accuracyforthedisulfideconnectivityproblemremains measured. The reason for this is simple—amino acidsthat flank half-cystines (disulfide-bonded cysteines) are quitedifferent from those that flank free cysteines (non-bondedcysteines) (6,7). In contrast, the residues that flank two incor-rectly paired half-cystines are quite similar to those that flankthe half-cystines in a disulfide bond. Two recent and remark-able papers based on different approaches (8,9) outperformearly attempts by Fariselli and co-workers (10,11). The VulloandFrasconimethod(9)usesrecursiveneuralnetworks(12)toscore undirected graphs that represent cysteine connectivity.The method of Zhao and co-workers (8) is based on recurrentpatterns of sequence separation between bonded half-cystines.Webserversthatallowonlinedisulfideconnectivitypredictionare available for Vullo/Frasconi (http://cassandra.dsi.unifi.it/cysteines) and, as a prototype, for Fariselli/Casadio (http://gpcr.biocomp.unibo.it/cgi/predictors/cys-cys/pred_dconcgi.cgi).Here, we describe a web server for disulfide connectivityprediction that implements our novel approach, which resultsin comparable and sometimes better than the state-of-the-artmethods (8,9). Algorithm details and performance of themethod are described previously by Ferre` and Clote (13).METHODSThe stand-alone program for disulfide connectivity prediction,implemented in our web server D

In search of peptide-based catalysts for asymmetric organic synthesis.

Abstract: The discovery of short peptide sequences that function as asymmetric catalysts for a variety of reactions is documented. The evolution of the project from an exercise in rational design to an endeavor that combines combinatorial screening with various mechanism-based experiments is presented. The specific development of catalysts for enantioselective acylation, phosphorylation, conjugate addition, and Morita-Baylis-Hillman reactions is described.