Publications

Recently developed chemical ligation protocols were elaborated for rapid N-terminal protein PEGylation. We introduce a PEG-salicylaldehyde ester and demonstrate its site-specific ligation to the N-terminus of the RNase S protein and to the therapeutic polypeptide PTH (1–34).

Peptoids are a family of synthetic oligomers composed of N-substituted glycine units. Along with other “foldamer” systems, peptoid oligomer sequences can be predictably designed to form a variety of stable secondary structures. It is not yet evident if foldamer design can be extended to reliably create tertiary structure features that mimic more complex biomolecular folds and functions. Computational modeling and prediction of peptoid conformations will likely play a critical role in enabling complex biomimetic designs. We introduce a computational approach to provide accurate conformational and energetic parameters for peptoid side chains needed for successful modeling and design. We find that peptoids can be described by a “rotamer” treatment, similar to that established for proteins, in which the peptoid side chains display rotational isomerism to populate discrete regions of the conformational landscape. Because of the insufficient number of solved peptoid structures, we have calculated the relative energies of side-chain conformational states to provide a backbone-dependent (BBD) rotamer library for a set of 54 different peptoid side chains. We evaluated two rotamer library development methods that employ quantum mechanics (QM) and/or molecular mechanics (MM) energy calculations to identify side-chain rotamers. We show by comparison to experimental peptoid structures that both methods provide an accurate prediction of peptoid side chain placements in folded peptoid oligomers and at protein interfaces. We have incorporated our peptoid rotamer libraries into ROSETTA, a molecular design package previously validated in the context of protein design and structure prediction.

A neuron is a basic physiological and computational unit of the brain. While much is known about the physiological properties of a neuron, its computational role is poorly understood. Here we propose to view a neuron as a signal processing device that represents the incoming streaming data matrix as a sparse vector of synaptic weights scaled by an outgoing sparse activity vector. Formally, a neuron minimizes a cost function comprising a cumulative squared representation error and regularization terms. We derive an online algorithm that minimizes such cost function by alternating between the minimization with respect to activity and with respect to synaptic weights. The steps of this algorithm reproduce well-known physiological properties of a neuron, such as weighted summation and leaky integration of synaptic inputs, as well as an Oja-like, but parameter-free, synaptic learning rule. Our theoretical framework makes several predictions, some of which can be verified by the existing data, others require further experiments. Such framework should allow modeling the function of neuronal circuits without necessarily measuring all the microscopic biophysical parameters, as well as facilitate the design of neuromorphic electronics.

In vertebrates, pluripotent pharyngeal mesoderm progenitors produce the cardiac precursors of the second heart field as well as the branchiomeric head muscles and associated stem cells. However, the mechanisms underlying the transition from multipotent progenitors to distinct muscle precursors remain obscured by the complexity of vertebrate embryos. Using Ciona intestinalis as a simple chordate model, we show that bipotent cardiopharyngeal progenitors are primed to activate both heart and pharyngeal muscle transcriptional programs, which progressively become restricted to corresponding precursors. The transcription factor COE (Collier/OLF/EBF) orchestrates the transition to pharyngeal muscle fate both by promoting an MRF-associated myogenic program in myoblasts and by maintaining an undifferentiated state in their sister cells through Notch-mediated lateral inhibition. The latter are stem cell-like muscle precursors that form most of the juvenile pharyngeal muscles. We discuss the implications of our findings for the development and evolution of the chordate cardiopharyngeal mesoderm.

Endospore-forming bacteria, with Bacillus subtilis being the prevalent model organism, belong to the phylum Firmicutes. Although the last common ancestor of all Firmicutes is likely to have been an endospore-forming species, not every lineage in the phylum has maintained the ability to produce endospores (hereafter, spores). In 1997, the release of the full genome sequence for B. subtilis strain 168 marked the beginning of the genomic era for the study of spore formation (sporulation). In this original genome sequence, 139 of the 4,100 protein-coding genes were annotated as sporulation genes. By the time a revised genome sequence with updated annotations was published in 2009, that number had increased significantly, especially since transcriptional profiling studies (transcriptomics) led to the identification of several genes expressed under the control of known sporulation transcription factors. Over the past decade, genome sequences for multiple spore-forming species have been released (including several strains in the Bacillus anthracis/Bacillus cereus group and many Clostridium species), and phylogenomic analyses have revealed many conserved sporulation genes. Parallel advances in transcriptomics led to the identification of small untranslated regulatory RNAs (sRNAs), including some that are expressed during sporulation. An extended array of -omics techniques, i.e., techniques designed to probe gene function on a genome-wide scale, such as proteomics, metabolomics, and high-throughput protein localization studies, have been implemented in microbiology. Combined with the use of new computational methods for predicting gene function and inferring regulatory relationships on a global scale, these -omics approaches are uncovering novel information about sporulation and a variety of other bacterial cell processes.

Chromatin is the driver of gene regulation, yet understanding the molecular interactions underlying chromatin factor combinatorial patterns (or the “chromatin codes”) remains a fundamental challenge in chromatin biology. Here we developed a global modeling framework that leverages chromatin profiling data to produce a systems-level view of the macromolecular complex of chromatin. Our model ultilizes maximum entropy modeling with regularization-based structure learning to statistically dissect dependencies between chromatin factors and produce an accurate probability distribution of chromatin code. Our unsupervised quantitative model, trained on genome-wide chromatin profiles of 73 histone marks and chromatin proteins from modENCODE, enabled making various data-driven inferences about chromatin profiles and interactions. We provided a highly accurate predictor of chromatin factor pairwise interactions validated by known experimental evidence, and for the first time enabled higher-order interaction prediction. Our predictions can thus help guide future experimental studies. The model can also serve as an inference engine for predicting unknown chromatin profiles — we demonstrated that with this approach we can leverage data from well-characterized cell types to help understand less-studied cell type or conditions.

Promoters and enhancers establish precise gene transcription patterns. The development of functional approaches for their identification in mammalian cells has been complicated by the size of these genomes. Here we report a high-throughput functional assay for directly identifying active promoter and enhancer elements called FIREWACh (Functional Identification of Regulatory Elements Within Accessible Chromatin), which we used to simultaneously assess over 80,000 DNA fragments derived from nucleosome-free regions within the chromatin of embryonic stem cells (ESCs) and identify 6,364 active regulatory elements. Many of these represent newly discovered ESC-specific enhancers, showing enriched binding-site motifs for ESC-specific transcription factors including SOX2, POU5F1 (OCT4) and KLF4. The application of FIREWACh to additional cultured cell types will facilitate functional annotation of the genome and expand our view of transcriptional network dynamics.

Recent results have shown the possibility of both reconstructing connectomes of small but biologically interesting circuits and extracting from these connectomes insights into their function. However, these reconstructions were heroic proof-of-concept experiments, requiring person-months of effort per neuron reconstructed, and will not scale to larger circuits, much less the brains of entire animals. In this paper we examine what will be required to generate and use substantially larger connectomes, finding five areas that need increased attention: firstly, imaging better suited to automatic reconstruction, with excellent z-resolution; secondly, automatic detection, validation, and measurement of synapses; thirdly, reconstruction methods that keep and use uncertainty metrics for every object, from initial images, through segmentation, reconstruction, and connectome queries; fourthly, processes that are fully incremental, so that the connectome may be used before it is fully complete; and finally, better tools for analysis of connectomes, once they are obtained.

Toward large-scale connectome reconstructions (PDF Download Available). Available from: https://www.researchgate.net/publication/260562044_Toward_large-scale_connectome_reconstructions [accessed Jul 26, 2017].

Predicting protein secondary structure is a fundamental problem in protein structure prediction. Here we present a new supervised generative stochastic network (GSN) based method to predict local secondary structure with deep hierarchical representations. GSN is a recently proposed deep learning technique (Bengio & Thibodeau-Laufer, 2013) to globally train deep generative model. We present the supervised extension of GSN, which learns a Markov chain to sample from a conditional distribution, and applied it to protein structure prediction. To scale the model to full-sized, high-dimensional data, like protein sequences with hundreds of amino acids, we introduce a convolutional architecture, which allows efficient learning across multiple layers of hierarchical representations. Our architecture uniquely focuses on predicting structured low-level labels informed with both low and high-level representations learned by the model. In our application this corresponds to labeling the secondary structure state of each amino-acid residue. We trained and tested the model on separate sets of non-homologous proteins sharing less than 30% sequence identity. Our model achieves 66.4% Q8 accuracy on the CB513 dataset, better than the previously reported best performance 64.9% (Wang et al., 2011) for this challenging secondary structure prediction problem.

The placement of neuronal cell bodies relative to the neuropile differs among species and brain areas. Cell bodies can be either embedded as in mammalian cortex or segregated as in invertebrates and some other vertebrate brain areas. Why are there such different arrangements? Here we suggest that the observed arrangements may simply be a reflection of wiring economy, a general principle that tends to reduce the total volume of the neuropile and hence the volume of the inclusions in it. Specifically, we suggest that the choice of embedded versus segregated arrangement is determined by which neuronal component — the cell body or the neurite connecting the cell body to the arbor — has a smaller volume. Our quantitative predictions are in agreement with existing and new measurements.