Improved prediction of protein secondary structure by use of sequence profiles and neural networks

Decoding functional proteome information in model organisms using protein language models

Protein language models have been tested and proved to be reliable when used on curated datasets but have not yet been applied to full proteomes. Accordingly, we tested how two different machine learning-based methods performed when decoding functional information from the proteomes of selected model organisms. We found that protein language models are more precise and informative than deep learning methods for all the species tested and across the three gene ontologies studied, and that they better recover functional information from transcriptomic experiments. The results obtained indicate that these language models are likely to be suitable for large-scale annotation and downstream analyses, and we recommend a guide for their use.

Deep Learning for Genomics: From Early Neural Nets to Modern Large Language Models

The data explosion driven by advancements in genomic research, such as high-throughput sequencing techniques, is constantly challenging conventional methods used in genomics. In parallel with the urgent demand for robust algorithms, deep learning has succeeded in various fields such as vision, speech, and text processing. Yet genomics entails unique challenges to deep learning, since we expect a superhuman intelligence that explores beyond our knowledge to interpret the genome from deep learning. A powerful deep learning model should rely on the insightful utilization of task-specific knowledge. In this paper, we briefly discuss the strengths of different deep learning models from a genomic perspective so as to fit each particular task with proper deep learning-based architecture, and we remark on practical considerations of developing deep learning architectures for genomics. We also provide a concise review of deep learning applications in various aspects of genomic research and point out current challenges and potential research directions for future genomics applications. We believe the collaborative use of ever-growing diverse data and the fast iteration of deep learning models will continue to contribute to the future of genomics.

Data-Driven Models for Predicting Intrinsically Disordered Protein Polymer Physics Directly from Composition or Sequence

The molecular-level understanding of intrinsically disordered proteins is challenging due to experimental characterization difficulties. Computational understanding of IDPs also requires fundamental advances, as the leading tools for predicting protein folding (e.g., AlphaFold), typically fail to describe the structural ensembles of IDPs. The focus of this paper is to 1) develop new representations for intrinsically disordered proteins and 2) pair these representations with classical machine learning and deep learning models to predict the radius of gyration and derived scaling exponent of IDPs. Here, we build a new physically-motivated feature called the bag of amino acid interactions representation, which encodes pairwise interactions explicitly into the representation. This feature essentially counts and weights all possible non-bonded interactions in a sequence and thus is, in principle, compatible with arbitrary sequence lengths. To see how well this new feature performs, both categorical and physically-motivated featurization techniques are tested on a computational dataset containing 10,000 sequences simulated at the coarse-grained level. The results indicate that this new feature outperforms the other purely categorical and physically-motivated features and possesses solid extrapolation capabilities. For future use, this feature can potentially provide physical insights into amino acid interactions, including their temperature dependence, and be applied to other protein spaces.

Theoretical and Data-Driven Approaches for Biomolecular Condensates

Biomolecular condensation processes are increasingly recognized as a fundamental mechanism that living cells use to organize biomolecules in time and space. These processes can lead to the formation of membraneless organelles that enable cells to perform distinct biochemical processes in controlled local environments, thereby supplying them with an additional degree of spatial control relative to that achieved by membrane-bound organelles. This fundamental importance of biomolecular condensation has motivated a quest to discover and understand the molecular mechanisms and determinants that drive and control this process. Within this molecular viewpoint, computational methods can provide a unique angle to studying biomolecular condensation processes by contributing the resolution and scale that are challenging to reach with experimental techniques alone. In this Review, we focus on three types of dry-lab approaches: theoretical methods, physics-driven simulations and data-driven machine learning methods. We review recent progress in using these tools for probing biomolecular condensation across all three fields and outline the key advantages and limitations of each of the approaches. We further discuss some of the key outstanding challenges that we foresee the community addressing next in order to develop a more complete picture of the molecular driving forces behind biomolecular condensation processes and their biological roles in health and disease.

Transcriptomic responses of the fast-growing bacterium Vibrio natriegens during cold-induced loss of culturability

Vibrio natriegens has massive biotechnological potential owing to its fast growth rate. However, this bacterium rapidly loses its culturability during low-temperature preservation (LTP), the reason for which is still unknown. To reveal the metabolic responses of V. natriegens during LTP, we analyzed and compared the transcriptome before and after 8 days of preservation at 4 or 25 °C (room-temperature preservation (RTP)) in liquid culture medium. Most genes exhibited significant transcriptional responses to LTP. Using gene set enrichment analysis, we compared the transcriptional responses of different V. natriegens Gene Ontology (GO) sets during LTP or RTP. The enrichment of the GO set "SOS response" during LTP, but not RTP, indicated the occurrence of DNA damage during LTP. The GO set "respiratory electron transport chain" was suppressed during LTP and RTP. Although the GO set "response to oxidative stress" was not significantly altered, we observed an increase in reactive oxygen species (ROS) during LTP, suggesting a relationship between ROS and cold-induced loss of culturability (CILC) in V. natriegens. The faster loss of culturability and accumulation of ROS in 20 mL compared to 100 mL of liquid culture medium further suggested a relationship between CILC and oxygen availability. Furthermore, we showed that the deletion of Na⁺-translocating NADH-ubiquinone oxidoreductase, but not type-II NADH dehydrogenase, accelerated CILC and increased intracellular ROS levels in V. natriegens. These findings will help to understand the cause of CILC which may lead to improving the stability of V. natriegens at low temperatures.

Pfeature: A Tool for Computing Wide Range of Protein Features and Building Prediction Models

In the last three decades, a wide range of protein features have been discovered to annotate a protein. Numerous attempts have been made to integrate these features in a software package/platform so that the user may compute a wide range of features from a single source. To complement the existing methods, we developed a method, Pfeature, for computing a wide range of protein features. Pfeature allows to compute more than 200,000 features required for predicting the overall function of a protein, residue-level annotation of a protein, and function of chemically modified peptides. It has six major modules, namely, composition, binary profiles, evolutionary information, structural features, patterns, and model building. Composition module facilitates to compute most of the existing compositional features, plus novel features. The binary profile of amino acid sequences allows to compute the fraction of each type of residue as well as its position. The evolutionary information module allows to compute evolutionary information of a protein in the form of a position-specific scoring matrix profile generated using Position-Specific Iterative Basic Local Alignment Search Tool (PSI-BLAST); fit for annotation of a protein and its residues. A structural module was developed for computing of structural features/descriptors from a tertiary structure of a protein. These features are suitable to predict the therapeutic potential of a protein containing non-natural or chemically modified residues. The model-building module allows to implement various machine learning techniques for developing classification and regression models as well as feature selection. Pfeature also allows the generation of overlapping patterns and features from a protein. A user-friendly Pfeature is available as a web server python library and stand-alone package.

ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

ProtTrans: Toward Understanding the Language of Life Through Self-Supervised Learning

Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive models (Transformer-XL, XLNet) and four auto-encoder models (BERT, Albert, Electra, T5) on data from UniRef and BFD containing up to 393 billion amino acids. The protein LMs (pLMs) were trained on the Summit supercomputer using 5616 GPUs and TPU Pod up-to 1024 cores. Dimensionality reduction revealed that the raw pLM-embeddings from unlabeled data captured some biophysical features of protein sequences. We validated the advantage of using the embeddings as exclusive input for several subsequent tasks: (1) a per-residue (per-token) prediction of protein secondary structure (3-state accuracy Q3=81%-87%); (2) per-protein (pooling) predictions of protein sub-cellular location (ten-state accuracy: Q10=81%) and membrane versus water-soluble (2-state accuracy Q2=91%). For secondary structure, the most informative embeddings (ProtT5) for the first time outperformed the state-of-the-art without multiple sequence alignments (MSAs) or evolutionary information thereby bypassing expensive database searches. Taken together, the results implied that pLMs learned some of the grammar of the language of life. All our models are available through https://github.com/agemagician/ProtTrans.

Different methods, techniques and their limitations in protein structure prediction: A review

Because of the increase in different types of diseases in human habitats, demands for designing various types of drugs are also increasing. Protein and its structure play a very important role in drug design. Therefore researchers from different areas like mathematics, medicines, and computer science are teaming up for getting better solutions in the said field. In this paper, we have discussed different methods of secondary and tertiary protein structure prediction (PSP), along with the limitations of different approaches. Different types of datasets used in PSP are also discussed here. This paper also tells about different performance measures to evaluate the prediction accuracy of PSP methods. Different software's/servers are available for download, which are used to find the protein structures for the input protein sequence. These softwares will also help to compare the performance of any new algorithm with other available methods. Details of those softwares are also mentioned in this paper.

Prediction of protein secondary structure based on an improved channel attention and multiscale convolution module

Prediction of the protein secondary structure is a key issue in protein science. Protein secondary structure prediction (PSSP) aims to construct a function that can map the amino acid sequence into the secondary structure so that the protein secondary structure can be obtained according to the amino acid sequence. Driven by deep learning, the prediction accuracy of the protein secondary structure has been greatly improved in recent years. To explore a new technique of PSSP, this study introduces the concept of an adversarial game into the prediction of the secondary structure, and a conditional generative adversarial network (GAN)-based prediction model is proposed. We introduce a new multiscale convolution module and an improved channel attention (ICA) module into the generator to generate the secondary structure, and then a discriminator is designed to conflict with the generator to learn the complicated features of proteins. Then, we propose a PSSP method based on the proposed multiscale convolution module and ICA module. The experimental results indicate that the conditional GAN-based protein secondary structure prediction (CGAN-PSSP) model is workable and worthy of further study because of the strong feature-learning ability of adversarial learning.

Lightweight ProteinUnet2 network for protein secondary structure prediction: a step towards proper evaluation

Background

The prediction of protein secondary structures is a crucial and significant step for ab initio tertiary structure prediction which delivers the information about proteins activity and functions. As the experimental methods are expensive and sometimes impossible, many SS predictors, mainly based on different machine learning methods have been proposed for many years. Currently, most of the top methods use evolutionary-based input features produced by PSSM and HHblits software, although quite recently the embeddings—the new description of protein sequences generated by language models (LM) have appeared that could be leveraged as input features. Apart from input features calculation, the top models usually need extensive computational resources for training and prediction and are barely possible to run on a regular PC. SS prediction as the imbalanced classification problem should not be judged by the commonly used Q3/Q8 metrics. Moreover, as the benchmark datasets are not random samples, the classical statistical null hypothesis testing based on the Neyman–Pearson approach is not appropriate.

Results

We present a lightweight deep network ProteinUnet2 for SS prediction which is based on U-Net convolutional architecture and evolutionary-based input features (from PSSM and HHblits) as well as SPOT-Contact features. Through an extensive evaluation study, we report the performance of ProteinUnet2 in comparison with top SS prediction methods based on evolutionary information (SAINT and SPOT-1D). We also propose a new statistical methodology for prediction performance assessment based on the significance from Fisher–Pitman permutation tests accompanied by practical significance measured by Cohen’s effect size.

Conclusions

Our results suggest that ProteinUnet2 architecture has much shorter training and inference times while maintaining results similar to SAINT and SPOT-1D predictors. Taking into account the relatively long times of calculating evolutionary-based features (from PSSM in particular), it would be worth conducting the predictive ability tests on embeddings as input features in the future. We strongly believe that our proposed here statistical methodology for the evaluation of SS prediction results will be adopted and used (and even expanded) by the research community.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04623-z.

Information quantity for secondary structure propensities of protein subsequences in the Protein Data Bank

Elucidating the principles of sequence-structure relationships of proteins is a long-standing issue in biology. The nature of a short segment of a protein is determined by both the subsequence of the segment itself and its environment. For example, a type of subsequence, the so-called chameleon sequences, can form different secondary structures depending on its environments. Chameleon sequences are considered to have a weak tendency to form a specific structure. Although many chameleon sequences have been identified, they are only a small part of all possible subsequences in the proteome. The strength of the tendency to take a specific structure for each subsequence has not been fully quantified. In this study, we comprehensively analyzed subsequences consisting of four to nine amino acid residues, or N-gram (4≤N≤9), observed in non-redundant sequences in the Protein Data Bank (PDB). Tendencies to form a specific structure in terms of the secondary structure and accessible surface area are quantified as information quantities for each N-gram. Although the majority of observed subsequences have low information quantity due to lack of samples in the current PDB, thousands of N-grams with strong tendencies, including known structural motifs, were found. In addition, machine learning partially predicted the tendency of unknown N-grams, and thus, this technique helps to extract knowledge from the limited number of samples in the PDB.

CSI-LSTM: a web server to predict protein secondary structure using bidirectional long short term memory and NMR chemical shifts

Protein secondary structure provides rich structural information, hence the description and understanding of protein structure relies heavily on it. Identification or prediction of secondary structures therefore plays an important role in protein research. In protein NMR studies, it is more convenient to predict secondary structures from chemical shifts as compared to the traditional determination methods based on inter-nuclear distances provided by NOESY experiment. In recent years, there was a significant improvement observed in deep neural networks, which had been applied in many research fields. Here we proposed a deep neural network based on bidirectional long short term memory (biLSTM) to predict protein 3-state secondary structure using NMR chemical shifts of backbone nuclei. While comparing with the existing methods the proposed method showed better prediction accuracy. Based on the proposed method, a web server has been built to provide protein secondary structure prediction service.

Discovering the Ultimate Limits of Protein Secondary Structure Prediction

Secondary structure prediction (SSP) of proteins is an important structural biology technique with many applications. There have been ~300 algorithms published in the past seven decades with fierce competition in accuracy. In the first 60 years, the accuracy of three-state SSP rose from ~56% to 81%; after that, it has long stayed at 81-86%. In the 1990s, the theoretical limit of three-state SSP accuracy had been estimated to be 88%. Thus, SSP is now generally considered not challenging or too challenging to improve. However, we found that the limit of three-state SSP might be underestimated. Besides, there is still much room for improving segment-based and eight-state SSPs, but the limits of these emerging topics have not been determined. This work performs large-scale sequence and structural analyses to estimate SSP accuracy limits and assess state-of-the-art SSP methods. The limit of three-state SSP is re-estimated to be ~92%, 4-5% higher than previously expected, indicating that SSP is still challenging. The estimated limit of eight-state SSP is 84-87%. Several proposals for improving future SSP algorithms are made based on our results. We hope that these findings will help move forward the development of SSP and all its applications.

Computational studies of anaplastic lymphoma kinase mutations reveal common mechanisms of oncogenic activation

High-risk tumors are genomically heterogeneous, harboring gene amplifications and mutations. The activation status of mutated proteins in cancer can profoundly impact disease progression, patient response, and drug sensitivity. Yet, outside of a few hotspot mutations, functional studies of clinically observed mutations are not commonly pursued. We report a combined experimental profiling and computational analysis of the effects of clinically observed and “test” mutations in the kinase domain of anaplastic lymphoma kinase (ALK), a known oncogenic driver in pediatric neuroblastoma. We find that the activation status of the mutated protein is a good indicator of the transforming ability in NIH 3T3 cells. We also report biophysical as well as data-driven models with predictive power to profile these mutant kinases in silico.

Kinases play important roles in diverse cellular processes, including signaling, differentiation, proliferation, and metabolism. They are frequently mutated in cancer and are the targets of a large number of specific inhibitors. Surveys of cancer genome atlases reveal that kinase domains, which consist of 300 amino acids, can harbor numerous (150 to 200) single-point mutations across different patients in the same disease. This preponderance of mutations—some activating, some silent—in a known target protein make clinical decisions for enrolling patients in drug trials challenging since the relevance of the target and its drug sensitivity often depend on the mutational status in a given patient. We show through computational studies using molecular dynamics (MD) as well as enhanced sampling simulations that the experimentally determined activation status of a mutated kinase can be predicted effectively by identifying a hydrogen bonding fingerprint in the activation loop and the αC-helix regions, despite the fact that mutations in cancer patients occur throughout the kinase domain. In our study, we find that the predictive power of MD is superior to a purely data-driven machine learning model involving biochemical features that we implemented, even though MD utilized far fewer features (in fact, just one) in an unsupervised setting. Moreover, the MD results provide key insights into convergent mechanisms of activation, primarily involving differential stabilization of a hydrogen bond network that engages residues of the activation loop and αC-helix in the active-like conformation (in >70% of the mutations studied, regardless of the location of the mutation).

A secondary structure-based position-specific scoring matrix applied to the improvement in protein secondary structure prediction

Protein secondary structure prediction (SSP) has a variety of applications; however, there has been relatively limited improvement in accuracy for years. With a vision of moving forward all related fields, we aimed to make a fundamental advance in SSP. There have been many admirable efforts made to improve the machine learning algorithm for SSP. This work thus took a step back by manipulating the input features. A secondary structure element-based position-specific scoring matrix (SSE-PSSM) is proposed, based on which a new set of machine learning features can be established. The feasibility of this new PSSM was evaluated by rigid independent tests with training and testing datasets sharing <25% sequence identities. In all experiments, the proposed PSSM outperformed the traditional amino acid PSSM. This new PSSM can be easily combined with the amino acid PSSM, and the improvement in accuracy was remarkable. Preliminary tests made by combining the SSE-PSSM and well-known SSP methods showed 2.0% and 5.2% average improvements in three- and eight-state SSP accuracies, respectively. If this PSSM can be integrated into state-of-the-art SSP methods, the overall accuracy of SSP may break the current restriction and eventually bring benefit to all research and applications where secondary structure prediction plays a vital role during development. To facilitate the application and integration of the SSE-PSSM with modern SSP methods, we have established a web server and standalone programs for generating SSE-PSSM available at http://10.life.nctu.edu.tw/SSE-PSSM.

A Multitask Deep-Learning Method for Predicting Membrane Associations and Secondary Structures of Proteins

Prediction of residue-level structural attributes and protein-level structural classes helps model protein tertiary structures and understand protein functions. Existing methods are either specialized on only one class of proteins or developed to predict only a specific type of residue-level attribute. In this work, we develop a new deep-learning method, named Membrane Association and Secondary Structure Predictor (MASSP), for accurately predicting both residue-level structural attributes (secondary structure, location, orientation, and topology) and protein-level structural classes (bitopic, α-helical, β-barrel, and soluble). MASSP integrates a multilayer two-dimensional convolutional neural network (2D-CNN) with a long short-term memory (LSTM) neural network into a multitasking framework. Our comparison shows that MASSP performs equally well or better than the state-of-the-art methods in predicting residue-level secondary structures, boundaries of transmembrane segments, and topology. Furthermore, it achieves outstanding accuracy in predicting protein-level structural classes. MASSP automatically distinguishes the structural classes of input sequences and identifies transmembrane segments and topologies if present, making it broadly applicable to different classes of proteins. In summary, MASSP's good performance and broad applicability make it well suited for annotating residue-level attributes and protein-level structural classes at the proteome scale.

PredictProtein - Predicting Protein Structure and Function for 29 Years

Since 1992 PredictProtein (https://predictprotein.org) is a one-stop online resource for protein sequence analysis with its main site hosted at the Luxembourg Centre for Systems Biomedicine (LCSB) and queried monthly by over 3,000 users in 2020. PredictProtein was the first Internet server for protein predictions. It pioneered combining evolutionary information and machine learning. Given a protein sequence as input, the server outputs multiple sequence alignments, predictions of protein structure in 1D and 2D (secondary structure, solvent accessibility, transmembrane segments, disordered regions, protein flexibility, and disulfide bridges) and predictions of protein function (functional effects of sequence variation or point mutations, Gene Ontology (GO) terms, subcellular localization, and protein-, RNA-, and DNA binding). PredictProtein's infrastructure has moved to the LCSB increasing throughput; the use of MMseqs2 sequence search reduced runtime five-fold (apparently without lowering performance of prediction methods); user interface elements improved usability, and new prediction methods were added. PredictProtein recently included predictions from deep learning embeddings (GO and secondary structure) and a method for the prediction of proteins and residues binding DNA, RNA, or other proteins. PredictProtein.org aspires to provide reliable predictions to computational and experimental biologists alike. All scripts and methods are freely available for offline execution in high-throughput settings.

Variable Length Character N-Gram Embedding of Protein Sequences for Secondary Structure Prediction

BACKGROUND

The prediction of a protein's secondary structure from its amino acid sequence is an essential step towards predicting its 3-D structure. The prediction performance improves by incorporating homologous multiple sequence alignment information. Since homologous details not available for all proteins. Therefore, it is necessary to predict the protein secondary structure from single sequences.

OBJECTIVE AND METHODS

Protein secondary structure predicted from their primary sequences using n-gram word embedding and deep recurrent neural network. Protein secondary structure depends on local and long-range neighbor residues in primary sequences. In the proposed work, the local contextual information of amino acid residues captures variable-length character n-gram words. An embedding vector represents these variable-length character n-gram words. Further, the bidirectional long short-term memory (Bi-LSTM) model is used to capture the long-range contexts by extracting the past and future residues information in primary sequences.

RESULTS

The proposed model evaluates on three public datasets ss.txt, RS126, and CASP9. The model shows the Q3 accuracy of 92.57%, 86.48%, and 89.66% for ss.txt, RS126, and CASP9.

CONCLUSION

The proposed model performance compares with state-of-the-art methods available in the literature. After a comparative analysis, it observed that the proposed model performs better than state-of-the-art methods.

Secondary Structure of the Novel Myosin Binding Domain WYR and Implications within Myosin Structure

Structural changes in the myosin II light meromyosin (LMM) that influence thick filament mechanical properties and muscle function are modulated by LMM-binding proteins. Flightin is an LMM-binding protein indispensable for the function of Drosophila indirect flight muscle (IFM). Flightin has a three-domain structure that includes WYR, a novel 52 aa domain conserved throughout Pancrustacea. In this study, we (i) test the hypothesis that WYR binds the LMM, (ii) characterize the secondary structure of WYR, and (iii) examine the structural impact WYR has on the LMM. Circular dichroism at 260-190 nm reveals a structural profile for WYR and supports an interaction between WYR and LMM. A WYR-LMM interaction is supported by co-sedimentation with a stoichiometry of ~2.4:1. The WYR-LMM interaction results in an overall increased coiled-coil content, while curtailing ɑ helical content. WYR is found to be composed of 15% turns, 31% antiparallel β, and 48% 'other' content. We propose a structural model of WYR consisting of an antiparallel β hairpin between Q92-K114 centered on an ASX or β turn around N102, with a G1 bulge at G117. The Drosophila LMM segment used, V1346-I1941, encompassing conserved skip residues 2-4, is found to possess a traditional helical profile but is interpreted as having <30% helical content by multiple methods of deconvolution. This low helicity may be affiliated with the dynamic behavior of the structure in solution or the inclusion of a known non-helical region in the C-terminus. Our results support the hypothesis that WYR binds the LMM and that this interaction brings about structural changes in the coiled-coil. These studies implicate flightin, via the WYR domain, for distinct shifts in LMM secondary structure that could influence the structural properties and stabilization of the thick filament, scaling to modulation of whole muscle function.

Protein Secondary Structure Prediction With a Reductive Deep Learning Method

Protein secondary structures have been identified as the links in the physical processes of primary sequences, typically random coils, folding into functional tertiary structures that enable proteins to involve a variety of biological events in life science. Therefore, an efficient protein secondary structure predictor is of importance especially when the structure of an amino acid sequence fragment is not solved by high-resolution experiments, such as X-ray crystallography, cryo-electron microscopy, and nuclear magnetic resonance spectroscopy, which are usually time consuming and expensive. In this paper, a reductive deep learning model MLPRNN has been proposed to predict either 3-state or 8-state protein secondary structures. The prediction accuracy by the MLPRNN on the publicly available benchmark CB513 data set is comparable with those by other state-of-the-art models. More importantly, taking into account the reductive architecture, MLPRNN could be a baseline for future developments.

Protein Secondary Structure Prediction Using Character Bi-gram Embedding and Bi-LSTM

Background:

Protein secondary structure is vital to predicting the tertiary structure, which is essential in deciding protein function and drug designing. Therefore, there is a high requirement of computational methods to predict secondary structure from their primary sequence. Protein primary sequences represented as a linear combination of twenty amino acid characters and contain the contextual information for secondary structure prediction.

Objective and Methods:

Protein secondary structure predicted from their primary sequences using a deep recurrent neural network. Protein secondary structure depends on local and long-range residues in primary sequences. In the proposed work, the local contextual information of amino acid residues captures with character n-gram. A dense embedding vector represents this local contextual information. Furthermore, the bidirectional long short-term memory (Bi-LSTM) model is used to capture the long-range contexts by extracting the past and future residues information in primary sequences.

Results:

The proposed deep recurrent architecture is evaluated for its efficacy for datasets, namely ss.txt, RS126, and CASP9. The model shows the Q3 accuracies of 88.45%, 83.48%, and 86.69% for ss.txt, RS126, and CASP9, respectively. The performance of the proposed model is also compared with other state-of-the-art methods available in the literature.

Conclusion:

After a comparative analysis, it was observed that the proposed model is performing better in comparison to state-of-art methods.

A Deep Convolutional Neural Network to Improve the Prediction of Protein Secondary Structure

Background:

Protein secondary structure prediction (PSSP) is a fundamental task in bioinformatics that is helpful for understanding the three-dimensional structure and biological function of proteins. Many neural network-based prediction methods have been developed for protein secondary structures. Deep learning and multiple features are two obvious means to improve prediction accuracy.

Objective:

To promote the development of PSSP, a deep convolutional neural network-based method is proposed to predict both the eight-state and three-state of protein secondary structure.

Methods:

In this model, sequence and evolutionary information of proteins are combined as multiple input features after preprocessing. A deep convolutional neural network with no pooling layer and connection layer is then constructed to predict the secondary structure of proteins. L2 regularization, batch normalization, and dropout techniques are employed to avoid over-fitting and obtain better prediction performance, and an improved cross-entropy is used as the loss function.

Results:

Our proposed model can obtain Q3 prediction results of 86.2%, 84.5%, 87.8%, and 84.7%, respectively, on CullPDB, CB513, CASP10 and CASP11 datasets, with corresponding Q8 prediction results of 74.1%, 70.5%, 74.9%, and 71.3%.

Conclusion:

We have proposed the DCNN-SS deep convolutional-network-based PSSP method, and experimental results show that DCNN-SS performs competitively with other methods.

Protein profiles: Biases and protocols

The use of evolutionary profiles to predict protein secondary structure, as well as other protein structural features, has been standard practice since the 1990s. Using profiles in the input of such predictors, in place or in addition to the sequence itself, leads to significantly more accurate predictions. While profiles can enhance structural signals, their role remains somewhat surprising as proteins do not use profiles when folding in vivo. Furthermore, the same sequence-based redundancy reduction protocols initially derived to train and evaluate sequence-based predictors, have been applied to train and evaluate profile-based predictors. This can lead to unfair comparisons since profiles may facilitate the bleeding of information between training and test sets. Here we use the extensively studied problem of secondary structure prediction to better evaluate the role of profiles and show that: (1) high levels of profile similarity between training and test proteins are observed when using standard sequence-based redundancy protocols; (2) the gain in accuracy for profile-based predictors, over sequence-based predictors, strongly relies on these high levels of profile similarity between training and test proteins; and (3) the overall accuracy of a profile-based predictor on a given protein dataset provides a biased measure when trying to estimate the actual accuracy of the predictor, or when comparing it to other predictors. We show, however, that this bias can be mitigated by implementing a new protocol (EVALpro) which evaluates the accuracy of profile-based predictors as a function of the profile similarity between training and test proteins. Such a protocol not only allows for a fair comparison of the predictors on equally hard or easy examples, but also reduces the impact of choosing a given similarity cutoff when selecting test proteins. The EVALpro program is available in the SCRATCH suite ( www.scratch.proteomics.ics.uci.edu) and can be downloaded at: www.download.igb.uci.edu/#evalpro.

Impact of protein dynamics on secondary structure prediction

Protein 3D structures support their biological functions. As the number of protein structures is negligible in regards to the number of available protein sequences, prediction methodologies relying only on protein sequences are essential tools. In this field, protein secondary structure prediction (PSSPs) is a mature area, and is considered to have reached a plateau. Nonetheless, proteins are highly dynamical macromolecules, a property that could impact the PSSP methods. Indeed, in a previous study, the stability of local protein conformations was evaluated demonstrating that some regions easily changed to another type of secondary structure. The protein sequences of this dataset were used by PSSPs and their results compared to molecular dynamics to investigate their potential impact on the quality of the secondary structure prediction. Interestingly, a direct link is observed between the quality of the prediction and the stability of the assignment to the secondary structure state. The more stable a local protein conformation is, the better the prediction will be. The secondary structure assignment not taken from the crystallized structures but from the conformations observed during the dynamics slightly increase the quality of the secondary structure prediction. These results show that evaluation of PSSPs can be done differently, but also that the notion of dynamics can be included in development of PSSPs and other approaches such as de novo approaches.

A simple strategy to enhance the speed of protein secondary structure prediction without sacrificing accuracy

The secondary structure prediction of proteins is a classic topic of computational structural biology with a variety of applications. During the past decade, the accuracy of prediction achieved by state-of-the-art algorithms has been >80%; meanwhile, the time cost of prediction increased rapidly because of the exponential growth of fundamental protein sequence data. Based on literature studies and preliminary observations on the relationships between the size/homology of the fundamental protein dataset and the speed/accuracy of predictions, we raised two hypotheses that might be helpful to determine the main influence factors of the efficiency of secondary structure prediction. Experimental results of size and homology reductions of the fundamental protein dataset supported those hypotheses. They revealed that shrinking the size of the dataset could substantially cut down the time cost of prediction with a slight decrease of accuracy, which could be increased on the contrary by homology reduction of the dataset. Moreover, the Shannon information entropy could be applied to explain how accuracy was influenced by the size and homology of the dataset. Based on these findings, we proposed that a proper combination of size and homology reductions of the protein dataset could speed up the secondary structure prediction while preserving the high accuracy of state-of-the-art algorithms. Testing the proposed strategy with the fundamental protein dataset of the year 2018 provided by the Universal Protein Resource, the speed of prediction was enhanced over 20 folds while all accuracy measures remained equivalently high. These findings are supposed helpful for improving the efficiency of researches and applications depending on the secondary structure prediction of proteins. To make future implementations of the proposed strategy easy, we have established a database of size and homology reduced protein datasets at http://10.life.nctu.edu.tw/UniRefNR.

The Future of Protein Secondary Structure Prediction Was Invented by Oleg Ptitsyn

When Oleg Ptitsyn and his group published the first secondary structure prediction for a protein sequence, they started a research field that is still active today. Oleg Ptitsyn combined fundamental rules of physics with human understanding of protein structures. Most followers in this field, however, use machine learning methods and aim at the highest (average) percentage correctly predicted residues in a set of proteins that were not used to train the prediction method. We show that one single method is unlikely to predict the secondary structure of all protein sequences, with the exception, perhaps, of future deep learning methods based on very large neural networks, and we suggest that some concepts pioneered by Oleg Ptitsyn and his group in the 70s of the previous century likely are today’s best way forward in the protein secondary structure prediction field.

Protein Secondary Structure Prediction: A Review of Progress and Directions

Background:

Over the last few decades, a search for the theory of protein folding has grown into a full-fledged research field at the intersection of biology, chemistry and informatics. Despite enormous effort, there are still open questions and challenges, like understanding the rules by which amino acid sequence determines protein secondary structure.

Objective:

In this review, we depict the progress of the prediction methods over the years and identify sources of improvement.

Methods:

The protein secondary structure prediction problem is described followed by the discussion on theoretical limitations, description of the commonly used data sets, features and a review of three generations of methods with the focus on the most recent advances. Additionally, methods with available online servers are assessed on the independent data set.

Results:

The state-of-the-art methods are currently reaching almost 88% for 3-class prediction and 76.5% for an 8-class prediction.

Conclusion:

This review summarizes recent advances and outlines further research directions.

The Order-Disorder Continuum: Linking Predictions of Protein Structure and Disorder through Molecular Simulation

Intrinsically disordered proteins (IDPs) and intrinsically disordered regions within proteins (IDRs) serve an increasingly expansive list of biological functions, including regulation of transcription and translation, protein phosphorylation, cellular signal transduction, as well as mechanical roles. The strong link between protein function and disorder motivates a deeper fundamental characterization of IDPs and IDRs for discovering new functions and relevant mechanisms. We review recent advances in experimental techniques that have improved identification of disordered regions in proteins. Yet, experimentally curated disorder information still does not currently scale to the level of experimentally determined structural information in folded protein databases, and disorder predictors rely on several different binary definitions of disorder. To link secondary structure prediction algorithms developed for folded proteins and protein disorder predictors, we conduct molecular dynamics simulations on representative proteins from the Protein Data Bank, comparing secondary structure and disorder predictions with simulation results. We find that structure predictor performance from neural networks can be leveraged for the identification of highly dynamic regions within molecules, linked to disorder. Low accuracy structure predictions suggest a lack of static structure for regions that disorder predictors fail to identify. While disorder databases continue to expand, secondary structure predictors and molecular simulations can improve disorder predictor performance, which aids discovery of novel functions of IDPs and IDRs. These observations provide a platform for the development of new, integrated structural databases and fusion of prediction tools toward protein disorder characterization in health and disease.

PSSP: Protein splice site prediction algorithm using Bayesian approach

This study aimed to introduce an algorithm and identify intein motif and blocks involved in protein splicing, and explore the underlying methods in the development of detection of protein motifs. Inteins are mobile protein splicing elements capable of self-splicing post-translationally. They exist in viruses and bacteriophage, notwithstanding this broad phylogenetic distribution, all inteins apportion common structural features. A method was developed to predict intein in a raw sequence, using a ranking and scoring scheme based on amino acid θ value tables. This method aided in the identification and assessment of patterns characterizing the intein sequences. New intein conserved properties are revealed and the known ones are described and localized. We have computed the θ value of each amino acid at block A positions +1 to +13, block B positions l+13 to l+26 and block G positions -7 to +1 for the three categories. The consensus amino acids thus found are listed at the end of each row. We gave statistics for the distance between the blocks, block A to B, block B to F, and block F to G with the average being 66.1, 294, and 10.2 amino acids, respectively. The actual blocks A, B, and G of the one intein found in vacuolar membrane ATPase subunit, a precursor protein, are ranked 1. The results indicate all of the block sequences that are found in nine proteins are ranked at top of the list. The intein sequence is used to search the databases for intein-like proteins. Understanding the functional, structural, and dynamical aspects of inteins is important for intein engineering and the betterment of intein database.

Modeling aspects of the language of life through transfer-learning protein sequences

BACKGROUND

Predicting protein function and structure from sequence is one important challenge for computational biology. For 26 years, most state-of-the-art approaches combined machine learning and evolutionary information. However, for some applications retrieving related proteins is becoming too time-consuming. Additionally, evolutionary information is less powerful for small families, e.g. for proteins from the Dark Proteome. Both these problems are addressed by the new methodology introduced here.

RESULTS

We introduced a novel way to represent protein sequences as continuous vectors (embeddings) by using the language model ELMo taken from natural language processing. By modeling protein sequences, ELMo effectively captured the biophysical properties of the language of life from unlabeled big data (UniRef50). We refer to these new embeddings as SeqVec (Sequence-to-Vector) and demonstrate their effectiveness by training simple neural networks for two different tasks. At the per-residue level, secondary structure (Q3 = 79% ± 1, Q8 = 68% ± 1) and regions with intrinsic disorder (MCC = 0.59 ± 0.03) were predicted significantly better than through one-hot encoding or through Word2vec-like approaches. At the per-protein level, subcellular localization was predicted in ten classes (Q10 = 68% ± 1) and membrane-bound were distinguished from water-soluble proteins (Q2 = 87% ± 1). Although SeqVec embeddings generated the best predictions from single sequences, no solution improved over the best existing method using evolutionary information. Nevertheless, our approach improved over some popular methods using evolutionary information and for some proteins even did beat the best. Thus, they prove to condense the underlying principles of protein sequences. Overall, the important novelty is speed: where the lightning-fast HHblits needed on average about two minutes to generate the evolutionary information for a target protein, SeqVec created embeddings on average in 0.03 s. As this speed-up is independent of the size of growing sequence databases, SeqVec provides a highly scalable approach for the analysis of big data in proteomics, i.e. microbiome or metaproteome analysis.

CONCLUSION

Transfer-learning succeeded to extract information from unlabeled sequence databases relevant for various protein prediction tasks. SeqVec modeled the language of life, namely the principles underlying protein sequences better than any features suggested by textbooks and prediction methods. The exception is evolutionary information, however, that information is not available on the level of a single sequence.

A Brief History of Protein Sorting Prediction

Ever since the signal hypothesis was proposed in 1971, the exact nature of signal peptides has been a focus point of research. The prediction of signal peptides and protein subcellular location from amino acid sequences has been an important problem in bioinformatics since the dawn of this research field, involving many statistical and machine learning technologies. In this review, we provide a historical account of how position-weight matrices, artificial neural networks, hidden Markov models, support vector machines and, lately, deep learning techniques have been used in the attempts to predict where proteins go. Because the secretory pathway was the first one to be studied both experimentally and through bioinformatics, our main focus is on the historical development of prediction methods for signal peptides that target proteins for secretion; prediction methods to identify targeting signals for other cellular compartments are treated in less detail.

Atomic Motif Recognition in (Bio)Polymers: Benchmarks From the Protein Data Bank

Rationalizing the structure and structure-property relations for complex materials such as polymers or biomolecules relies heavily on the identification of local atomic motifs, e.g., hydrogen bonds and secondary structure patterns, that are seen as building blocks of more complex supramolecular and mesoscopic structures. Over the past few decades, several automated procedures have been developed to identify these motifs in proteins given the atomic structure. Being based on a very precise understanding of the specific interactions, these heuristic criteria formulate the question in a way that implies the answer, by defining a list of motifs based on those that are known to be naturally occurring. This makes them less likely to identify unexpected phenomena, such as the occurrence of recurrent motifs in disordered segments of proteins, and less suitable to be applied to different polymers whose structure is not driven by hydrogen bonds, or even to polypeptides when appearing in unusual, non-biological conditions. Here we discuss how unsupervised machine learning schemes can be used to recognize patterns based exclusively on the frequency with which different motifs occur, taking high-resolution structures from the Protein Data Bank as benchmarks. We first discuss the application of a density-based motif recognition scheme in combination with traditional representations of protein structure (namely, interatomic distances and backbone dihedrals). Then, we proceed one step further toward an entirely unbiased scheme by using as input a structural representation based on the atomic density and by employing supervised classification to objectively assess the role played by the representation in determining the nature of atomic-scale patterns.

Relative abundance and molecular evolution of Lake Sinai Virus (Sinaivirus) clades

Lake Sinai Viruses (Sinaivirus) are commonly detected in honey bees (Apis mellifera) but no disease phenotypes or fitness consequences have yet been demonstrated. This viral group is genetically diverse, lacks obvious geographic structure, and multiple lineages can co-infect individual bees. While phylogenetic analyses have been performed, the molecular evolution of LSV has not been studied extensively. Here, I use LSV isolates from GenBank as well as contigs assembled from honey bee Sequence Read Archive (SRA) accessions to better understand the evolutionary history of these viruses. For each ORF, substitution rate variation, codon usage, and tests of positive selection were evaluated. Outlier regions of high or low diversity were sought with sliding window analysis and the role of recombination in creating LSV diversity was explored. Phylogenetic analysis consistently identified two large clusters of sequences that correspond to the current LSV1 and LSV2 nomenclature, however lineages sister to LSV1 were the most frequently detected in honey bee SRA accessions. Different expression levels among ORFs suggested the occurrence of subgenomic transcripts. ORF1 and RNA-dependent RNA polymerase had higher evolutionary rates than the capsid and ORF4. A hypervariable region of the ORF1 protein-coding sequence was identified that had reduced selective constraint, but a site-based model of positive selection was not significantly more likely than a neutral model for any ORF. The only significant recombination signals detected between LSV1 and LSV2 initiated within this hypervariable region, but assumptions of the test (single-frame coding and independence of substitution rate by site) were violated. LSV codon usage differed strikingly from that of honey bees and other common honey-bee viruses, suggesting LSV is not strongly co-evolved with that host. LSV codon usage was significantly correlated with that of Varroa destructor, however, despite the relatively weak codon bias exhibited by the latter. While codon usage between the LSV1 and LSV2 clusters was similar for three ORFs, ORF4 codon usage was uncorrelated between these clades, implying rapid divergence of codon use for this ORF only. Phylogenetic placement and relative abundance of LSV isolates reconstructed from SRA accessions suggest that detection biases may be over-representing LSV1 and LSV2 in public databases relative to their sister lineages.

Computational Prediction of Secondary and Supersecondary Structures from Protein Sequences

Many new methods for the sequence-based prediction of the secondary and supersecondary structures have been developed over the last several years. These and older sequence-based predictors are widely applied for the characterization and prediction of protein structure and function. These efforts have produced countless accurate predictors, many of which rely on state-of-the-art machine learning models and evolutionary information generated from multiple sequence alignments. We describe and motivate both types of predictions. We introduce concepts related to the annotation and computational prediction of the three-state and eight-state secondary structure as well as several types of supersecondary structures, such as β hairpins, coiled coils, and α-turn-α motifs. We review 34 predictors focusing on recent tools and provide detailed information for a selected set of 14 secondary structure and 3 supersecondary structure predictors. We conclude with several practical notes for the end users of these predictive methods.

Improving prediction of protein secondary structure, backbone angles, solvent accessibility and contact numbers by using predicted contact maps and an ensemble of recurrent and residual convolutional neural networks

Motivation

Sequence-based prediction of one dimensional structural properties of proteins has been a long-standing subproblem of protein structure prediction. Recently, prediction accuracy has been significantly improved due to the rapid expansion of protein sequence and structure libraries and advances in deep learning techniques, such as residual convolutional networks (ResNets) and Long-Short-Term Memory Cells in Bidirectional Recurrent Neural Networks (LSTM-BRNNs). Here we leverage an ensemble of LSTM-BRNN and ResNet models, together with predicted residue-residue contact maps, to continue the push towards the attainable limit of prediction for 3- and 8-state secondary structure, backbone angles (θ, τ, ϕ and ψ), half-sphere exposure, contact numbers and solvent accessible surface area (ASA).

Results

The new method, named SPOT-1D, achieves similar, high performance on a large validation set and test set (≈1000 proteins in each set), suggesting robust performance for unseen data. For the large test set, it achieves 87% and 77% in 3- and 8-state secondary structure prediction and 0.82 and 0.86 in correlation coefficients between predicted and measured ASA and contact numbers, respectively. Comparison to current state-of-the-art techniques reveals substantial improvement in secondary structure and backbone angle prediction. In particular, 44% of 40-residue fragment structures constructed from predicted backbone Cα-based θ and τ angles are less than 6 Å root-mean-squared-distance from their native conformations, nearly 20% better than the next best. The method is expected to be useful for advancing protein structure and function prediction.

Availability and implementation

SPOT-1D and its data is available at: http://sparks-lab.org/.

Supplementary information

Supplementary data are available at Bioinformatics online.

Seeking allosteric networks in PDZ domains

Ever since Ranganathan and coworkers subjected the covariation of amino acid residues in the postsynaptic density-95/Discs large/Zonula occludens 1 (PDZ) domain family to a statistical correlation analysis, PDZ domains have represented a paradigmatic family to explore single domain protein allostery. Nevertheless, several theoretical and experimental studies in the past two decades have contributed contradicting results with regard to structural localization of the allosteric networks, or even questioned their actual existence in PDZ domains. In this review, we first describe theoretical and experimental approaches that were used to probe the energetic network(s) in PDZ domains. We then compare the proposed networks for two well-studied PDZ domains namely the third PDZ domain from PSD-95 and the second PDZ domain from PTP-BL. Our analysis highlights the contradiction between the different methods and calls for additional work to better understand these allosteric phenomena.

Protein Solvent-Accessibility Prediction by a Stacked Deep Bidirectional Recurrent Neural Network

Residue solvent accessibility is closely related to the spatial arrangement and packing of residues. Predicting the solvent accessibility of a protein is an important step to understand its structure and function. In this work, we present a deep learning method to predict residue solvent accessibility, which is based on a stacked deep bidirectional recurrent neural network applied to sequence profiles. To capture more long-range sequence information, a merging operator was proposed when bidirectional information from hidden nodes was merged for outputs. Three types of merging operators were used in our improved model, with a long short-term memory network performing as a hidden computing node. The trained database was constructed from 7361 proteins extracted from the PISCES server using a cut-off of 25% sequence identity. Sequence-derived features including position-specific scoring matrix, physical properties, physicochemical characteristics, conservation score and protein coding were used to represent a residue. Using this method, predictive values of continuous relative solvent-accessible area were obtained, and then, these values were transformed into binary states with predefined thresholds. Our experimental results showed that our deep learning method improved prediction quality relative to current methods, with mean absolute error and Pearson’s correlation coefficient values of 8.8% and 74.8%, respectively, on the CB502 dataset and 8.2% and 78%, respectively, on the Manesh215 dataset.

Sixty-five years of the long march in protein secondary structure prediction: the final stretch?

Protein secondary structure prediction began in 1951 when Pauling and Corey predicted helical and sheet conformations for protein polypeptide backbone even before the first protein structure was determined. Sixty-five years later, powerful new methods breathe new life into this field. The highest three-state accuracy without relying on structure templates is now at 82-84%, a number unthinkable just a few years ago. These improvements came from increasingly larger databases of protein sequences and structures for training, the use of template secondary structure information and more powerful deep learning techniques. As we are approaching to the theoretical limit of three-state prediction (88-90%), alternative to secondary structure prediction (prediction of backbone torsion angles and Cα-atom-based angles and torsion angles) not only has more room for further improvement but also allows direct prediction of three-dimensional fragment structures with constantly improved accuracy. About 20% of all 40-residue fragments in a database of 1199 non-redundant proteins have <6 Å root-mean-squared distance from the native conformations by SPIDER2. More powerful deep learning methods with improved capability of capturing long-range interactions begin to emerge as the next generation of techniques for secondary structure prediction. The time has come to finish off the final stretch of the long march towards protein secondary structure prediction.

Accurate Prediction of One-Dimensional Protein Structure Features Using SPINE-X

Accurate prediction of protein secondary structure and other one-dimensional structure features is essential for accurate sequence alignment, three-dimensional structure modeling, and function prediction. SPINE-X is a software package to predict secondary structure as well as accessible surface area and dihedral angles ϕ and ψ. For secondary structure SPINE-X achieves an accuracy of between 81 and 84 % depending on the dataset and choice of tests. The Pearson correlation coefficient for accessible surface area prediction is 0.75 and the mean absolute error from the ϕ and ψ dihedral angles are 20^∘ and 33^∘, respectively. The source code and a Linux executables for SPINE-X are available from Research and Information Systems at http://mamiris.com .

Protein function in precision medicine: deep understanding with machine learning

Precision medicine and personalized health efforts propose leveraging complex molecular, medical and family history, along with other types of personal data toward better life. We argue that this ambitious objective will require advanced and specialized machine learning solutions. Simply skimming some low-hanging results off the data wealth might have limited potential. Instead, we need to better understand all parts of the system to define medically relevant causes and effects: how do particular sequence variants affect particular proteins and pathways? How do these effects, in turn, cause the health or disease-related phenotype? Toward this end, deeper understanding will not simply diffuse from deeper machine learning, but from more explicit focus on understanding protein function, context-specific protein interaction networks, and impact of variation on both.

Comparison of Algorithms for Prediction of Protein Structural Features from Evolutionary Data

Proteins have many functions and predicting these is still one of the major challenges in theoretical biophysics and bioinformatics. Foremost amongst these functions is the need to fold correctly thereby allowing the other genetically dictated tasks that the protein has to carry out to proceed efficiently. In this work, some earlier algorithms for predicting protein domain folds are revisited and they are compared with more recently developed methods. In dealing with intractable problems such as fold prediction, when different algorithms show convergence onto the same result there is every reason to take all algorithms into account such that a consensus result can be arrived at. In this work it is shown that the application of different algorithms in protein structure prediction leads to results that do not converge as such but rather they collude in a striking and useful way that has never been considered before.

Glycogen: Biosynthesis and Regulation

Glycogen accumulation occurs in Escherichia coli and Salmonella enterica serovar Typhimurium as well as in many other bacteria. Glycogen will be formed when there is an excess of carbon under conditions in which growth is limited because of the lack of a growth nutrient, e.g., a nitrogen source. This review describes the enzymatic reactions involved in glycogen synthesis and the allosteric regulation of the first enzyme, ADP-glucose pyrophosphorylase. The properties of the enzymes involved in glycogen synthesis, ADP-glucose pyrophosphorylase, glycogen synthase, and branching enzyme are also characterized. The data describing the genetic regulation of the glycogen synthesis are also presented. An alternate pathway for glycogen synthesis in mycobacteria is also described.