Quantifying optimal accuracy of local primary sequence bioinformatics methods

MOTIVATION: Traditional bioinformatics methods scan primary sequences for local patterns. It is important to assess how accurate local primary sequence methods can be. RESULTS: We study the problem of donor pre-mRNA splice site recognition, where the sequence overlaps between real and decoy datasets can be quantified, exposing the intrinsic limitations of the performance of local primary sequence methods. We assess the accuracy of primary sequence methods generally by studying how they scale with dataset size and demonstrate that our new primary sequence ranking methods have superior performance.     

A homology identification method that combines protein sequence and structure information.

A new method is presented for identifying distantly related homologous proteins that are unrecognizable by conventional sequence comparison methods. The method combines information about functionally conserved sequence patterns with information about structure context. This information is encoded in stochastic discrete state-space models (DSMs) that comprise a new family of hidden Markov models. The new models are called sequence-pattern-embedded DSMs (pDSMs). This method can identify distantly related protein family members with a high sensitivity and specificity. The method is illustrated with trypsin-like serine proteases and globins. The strategy for building pDSMs is presented. The method has been validated using carefully constructed positive and negative control sets. In addition to the ability to recognize remote homologs, pDSM sequence analysis predicts secondary structures with higher sensitivity, specificity, and Q3 accuracy than DSM analysis, which omits information about conserved sequence patterns. The identification of trypsin-like serine proteases in new genomes is discussed.     

Fast model-based protein homology detection without alignment

MOTIVATION: As more genomes are sequenced, the demand for fast gene classification techniques is increasing. To analyze a newly sequenced genome, first the genes are identified and translated into amino acid sequences which are then classified into structural or functional classes. The best-performing protein classification methods are based on protein homology detection using sequence alignment methods. Alignment methods have recently been enhanced by discriminative methods like support vector machines (SVMs) as well as by position-specific scoring matrices (PSSM) as obtained from PSI-BLAST. However, alignment methods are time consuming if a new sequence must be compared to many known sequences-the same holds for SVMs. Even more time consuming is to construct a PSSM for the new sequence. The best-performing methods would take about 25 days on present-day computers to classify the sequences of a new genome (20,000 genes) as belonging to just one specific class--however, there are hundreds of classes. Another shortcoming of alignment algorithms is that they do not build a model of the positive class but measure the mutual distance between sequences or profiles. Only multiple alignments and hidden Markov models are popular classification methods which build a model of the positive class but they show low classification performance. The advantage of a model is that it can be analyzed for chemical properties common to the class members to obtain new insights into protein function and structure. We propose a fast model-based recurrent neural network for protein homology detection, the 'Long Short-Term Memory' (LSTM). LSTM automatically extracts indicative patterns for the positive class, but in contrast to profile methods it also extracts negative patterns and uses correlations between all detected patterns for classification. LSTM is capable to automatically extract useful local and global sequence statistics like hydrophobicity, polarity, volume, polarizability and combine them with a pattern. These properties make LSTM complementary to alignment-based approaches as it does not use predefined similarity measures like BLOSUM or PAM matrices. RESULTS: We have applied LSTM to a well known benchmark for remote protein homology detection, where a protein must be classified as belonging to a SCOP superfamily. LSTM reaches state-of-the-art classification performance but is considerably faster for classification than other approaches with comparable classification performance. LSTM is five orders of magnitude faster than methods which perform slightly better in classification and two orders of magnitude faster than the fastest SVM-based approaches (which, however, have lower classification performance than LSTM). Only PSI-BLAST and HMM-based methods show comparable time complexity as LSTM, but they cannot compete with LSTM in classification performance. To test the modeling capabilities of LSTM, we applied LSTM to PROSITE classes and interpreted the extracted patterns. In 8 out of 15 classes, LSTM automatically extracted the PROSITE motif. In the remaining 7 cases alternative motifs are generated which give better classification results on average than the PROSITE motifs. AVAILABILITY: The LSTM algorithm is available from http://www.bioinf.jku.at/software/LSTM_protein/.     

A model for statistical significance of local similarities in structure

Structural biology can provide three-dimensional structures for proteins of unknown function. When sequence or structure comparisons fail to suggest a function, insights can come from discovery of functionally important local structural patterns. Existing methods to detect such patterns lack rigorous statistics needed for widespread application. Here, we derive a formula to calculate statistical significance of the root-mean-square deviation between atoms in such patterns. When combined with a database search method, our statistics permit true functional or structural patterns in different folds to be discerned from noise. The approach is highly complementary to fold comparison for providing functional clues for new structures, and is key for the detection of recurrences of any new pattern.     

Mouse oncogene protein 24p3 is a member of the lipocalin protein family.

Rigorous new methods of protein sequence analysis have been applied to the lipocalins, a diverse family of ligand binding proteins. Using three conserved sequence motifs to search for similar patterns in a large sequence database, the size and composition of this protein family have been defined in an automatic and objective way. It has allowed the identification of an existing sequence, mouse 24p3 protein, as a lipocalin and the possible rejection of other putative members from this protein family. On the basis of this newly discovered homology, a possible function for mouse 24p3 protein is proposed.     

AutoSCOP: automated prediction of SCOP classifications using unique pattern-class mappings

MOTIVATION: The sequence patterns contained in the available motif and hidden Markov model (HMM) databases are a valuable source of information for protein sequence annotation. For structure prediction and fold recognition purposes, we computed mappings from such pattern databases to the protein domain hierarchy given by the ASTRAL compendium and applied them to the prediction of SCOP classifications. Our aim is to make highly confident predictions also for non-trivial cases if possible and abstain from a prediction otherwise, and thus to provide a method that can be used as a first step in a pipeline of prediction methods. We describe two successful examples for such pipelines. With the AutoSCOP approach, it is possible to make predictions in a large-scale manner for many domains of the available sequences in the well-known protein sequence databases. RESULTS: AutoSCOP computes unique sequence patterns and pattern combinations for SCOP classifications. For instance, we assign a SCOP superfamily to a pattern found in its members whenever the pattern does not occur in any other SCOP superfamily. Especially on the fold and superfamily level, our method achieves both high sensitivity (above 93%) and high specificity (above 98%) on the difference set between two ASTRAL versions, due to being able to abstain from unreliable predictions. Further, on a harder test set filtered at low sequence identity, the combination with profile-profile alignments improves accuracy and performs comparably even to structure alignment methods. Integrating our method with structure alignment, we are able to achieve an accuracy of 99% on SCOP fold classifications on this set. In an analysis of false assignments of domains from new folds/superfamilies/families to existing SCOP classifications, AutoSCOP correctly abstains for more than 70% of the domains belonging to new folds and superfamilies, and more than 80% of the domains belonging to new families. These findings show that our approach is a useful additional filter for SCOP classification prediction of protein domains in combination with well-known methods such as profile-profile alignment. AVAILABILITY: A web server where users can input their domain sequences is available at http://www.bio.ifi.lmu.de/autoscop.     

3MATRIX and 3MOTIF: a protein structure visualization system for conserved sequence motifs

Computational methods such as sequence alignment and motif construction are useful in grouping related proteins into families, as well as helping to annotate new proteins of unknown function. These methods identify conserved amino acids in protein sequences, but cannot determine the specific functional or structural roles of conserved amino acids without additional study. In this work, we present 3MATRIX (http://3matrix.stanford.edu) and 3MOTIF (http://3motif.stanford.edu), a web-based sequence motif visualization system that displays sequence motif information in its appropriate three-dimensional (3D) context. This system is flexible in that users can enter sequences, keywords, structures or sequence motifs to generate visualizations. In 3MOTIF, users can search using discrete sequence motifs such as PROSITE patterns, eMOTIFs, or any other regular expression-like motif. Similarly, 3MATRIX accepts an eMATRIX position-specific scoring matrix, or will convert a multiple sequence alignment block into an eMATRIX for visualization. Each query motif is used to search the protein structure database for matches, in which the motif is then visually highlighted in three dimensions. Important properties of motifs such as sequence conservation and solvent accessible surface area are also displayed in the visualizations, using carefully chosen color shading schemes.     

Structure-based function inference using protein family-specific fingerprints

We describe a method to assign a protein structure to a functional family using family-specific fingerprints. Fingerprints represent amino acid packing patterns that occur in most members of a family but are rare in the background, a nonredundant subset of PDB; their information is additional to sequence alignments, sequence patterns, structural superposition, and active-site templates. Fingerprints were derived for 120 families in SCOP using Frequent Subgraph Mining. For a new structure, all occurrences of these family-specific fingerprints may be found by a fast algorithm for subgraph isomorphism; the structure can then be assigned to a family with a confidence value derived from the number of fingerprints found and their distribution in background proteins. In validation experiments, we infer the function of new members added to SCOP families and we discriminate between structurally similar, but functionally divergent TIM barrel families. We then apply our method to predict function for several structural genomics proteins, including orphan structures. Some predictions have been corroborated by other computational methods and some validated by subsequent functional characterization.     

Sequence patterns associated with disordered regions in proteins

The relationship between amino acid sequence and intrinsic disorder in proteins is investigated. Two databases, one of disordered proteins and the other of globular proteins, are analyzed and compared in order to extract simple sequence patterns of a few amino acids or amino acid properties that characterize disordered segments. It is found that a number of reliable, nonrandom associations exists. In particular, two types of patterns appear to be recurrent: a proline-rich pattern and a (positively or negatively) charged pattern. These results indicate that local sequence information can determine disordered regions in proteins. The derived patterns provide some insights into the physical reasons for disordered structures. They should also be helpful in improving currently available prediction methods.     

'Multifrequency' location and clustering of sequence patterns from proteins

In previous work, we have shown that a set of characteristics,defined as (code frequency) pairs, can be derived from a proteinfamily by the use of a signal-processing method. This methodenables the location and extraction of sequence patterns bytaking into account each (code frequency) pair individually.In the present paper, we propose to extend this method in orderto detect and visualize patterns by taking into account severalpairs simultaneously. Two ‘multifrequency’ methodsare described. The first one is based on a rewriting of thesequences with new symbols which summarize the frequency information.The second method is based on a clustering of the patterns associatedwith each pair. Both methods lead to the definition of significantconsensus sequences. Some results obtained with calcium-bindingproteins and serine proteases are also discussed. Received on March 6, 1990; accepted on September 24, 1990     

Molecular evidence on plant divergence times

Estimation of divergence times from sequence data has become increasingly feasible in recent years. Conflicts between fossil evidence and molecular dates have sparked the development of new methods for inferring divergence times, further encouraging these efforts. In this paper, available methods for estimating divergence times are reviewed, especially those geared toward handling the widespread variation in rates of molecular evolution observed among lineages. The assumptions, strengths, and weaknesses of local clock, Bayesian, and rate smoothing methods are described. The rapidly growing literature applying these methods to key divergence times in plant evolutionary history is also reviewed. These include the crown group ages of green plants, land plants, seed plants, angiosperms, and major subclades of angiosperms. Finally, attempts to infer divergence times are described in the context of two very different temporal settings: recent adaptive radiations and much more ancient biogeographic patterns.     

N-糖基化位点鉴定方法和非经典N-糖基化序列

蛋白质的N-糖基化修饰是生物体调控蛋白质在组织和细胞中的定位、功能、活性、寿命和多样性的一种普遍的翻译后方式。N-糖基化位点是理解糖链功能的重要前提之一。应用新的糖蛋白、糖肽富集技术和质谱技术,科学家们在不同组织中完成了对N-糖基化位点的鉴定。此外,不同于经典三联子的N-糖基化序列的发现使人们对N-糖基化过程的认识向纵深发展。     

Analysis of loop boundaries using different local structure assignment methods

Loops connect regular secondary structures. In many instances, they are known to play important biological roles. Analysis and prediction of loop conformations depend directly on the definition of repetitive structures. Nonetheless, the secondary structure assignment methods (SSAMs) often lead to divergent assignments. In this study, we analyzed, both structure and sequence point of views, how the divergence between different SSAMs affect boundary definitions of loops connecting regular secondary structures. The analysis of SSAMs underlines that no clear consensus between the different SSAMs can be easily found. Because these latter greatly influence the loop boundary definitions, important variations are indeed observed, that is, capping positions are shifted between different SSAMs. On the other hand, our results show that the sequence information in these capping regions are more stable than expected, and, classical and equivalent sequence patterns were found for most of the SSAMs. This is, to our knowledge, the most exhaustive survey in this field as (i) various databank have been used leading to similar results without implication of protein redundancy and (ii) the first time various SSAMs have been used. This work hence gives new insights into the difficult question of assignment of repetitive structures and addresses the issue of loop boundaries definition. Although SSAMs give very different local structure assignments capping sequence patterns remain efficiently stable.     

Identification of functionally conserved residues with the use of entropy-variability plots

We introduce sequence entropy-variability plots as a method of analyzing families of protein sequences, and demonstrate this for three well-known sequence families: globins, ras-like proteins, and serine-proteases. The location of an aligned residue position in the entropy-variability plot correlates with structural characteristics, and with known facts about the roles of individual amino acids in the function of these proteins. The large numbers of known sequences in these families allowed us to introduce new filtering methods for variability patterns. The results are discussed in terms of a simple evolutionary model for functional proteins.     

New deep learning-based methods for visualizing ecosystem properties using environmental DNA metabarcoding data

Environmental DNA (eDNA) metabarcoding provides an efficient approach for documenting biodiversity patterns in marine and terrestrial ecosystems. The complexity of these data prevents current methods from extracting and analyzing all the relevant ecological information they contain, and new methods may provide better dimensionality reduction and clustering. Here we present two new deep learning-based methods that combine different types of neural networks (NNs) to ordinate eDNA samples and visualize ecosystem properties in a two-dimensional space: the first is based on variational autoencoders and the second on deep metric learning. The strength of our new methods lies in the combination of two inputs: the number of sequences found for each molecular operational taxonomic unit (MOTU) detected and their corresponding nucleotide sequence. Using three different datasets, we show that our methods accurately represent several biodiversity indicators in a two-dimensional latent space: MOTU richness per sample, sequence α-diversity per sample, Jaccard's and sequence β-diversity between samples. We show that our nonlinear methods are better at extracting features from eDNA datasets while avoiding the major biases associated with eDNA. Our methods outperform traditional dimension reduction methods such as Principal Component Analysis, t-distributed Stochastic Neighbour Embedding, Nonmetric Multidimensional Scaling and Uniform Manifold Approximation and Projection for dimension reduction. Our results suggest that NNs provide a more efficient way of extracting structure from eDNA metabarcoding data, thereby improving their ecological interpretation and thus biodiversity monitoring.     

Haplotype Parsing

While the shared consensus genetic sequence of our species contains a great deal of information about our common biology, there is also much to be learned from the subtle genetic variations across our species. These variations are believed to be generally of little or no direct functional significance and predominantly reflect the chance accumulation of small genetic changes since our emergence as a species. Therefore, they carry little useful information when observed in a single individual. When tallied across a whole population though, these chance mutations can teach us a great deal about our evolutionary history and the patterns of inheritance in particular individuals. In particular, frequently observed patterns of single nucleotide polymorphisms (SNPs) in a population can identify segments of chromosome that have been passed down largely intact through long stretches of our evolution. Finding these frequently conserved chromosomal segments, or haplotypes, and developing methods to identify haplotype patterns in particular individuals, will in turn help us to identify those particular segments that carry genetic factors influencing risk for many common human diseases. To make the best use of this data, we will need to develop new models for the encoding of information in genome variations--the "language of genetic variation"--and new algorithms for fitting datasets to those models. This article surveys past work by the author and colleagues on this problem, utilising computational methods for locating frequent patterns in haploid sequence data, and "parsing" sequences so as to optimally explain them given the knowledge of the general population structure. The author's recent work in this area has been compiled into a set of computational tools available at http://www-2.cs.cmu.edu/~russells/software/hapmotif.html.     

Amino acid sequence templates derived from recurrent turn motifs in proteins: critical evaluation of their predictive power

Amino acid sequence patterns suggested to characterize specific recurrent turn conformation in protein are tested as to their predictive power in a database containing 75 proteins of known structure. Many of these patterns are found to be associated with local structures that differ from the motifs originally used to derive them. It is therefore concluded that, while they could be useful for improving predictions made by other methods, their stand-alone predictive power is poor. The issue of deriving and validating consensus sequence patterns for use in protein structure prediction is raised.     

Chaos game representation of gene structure.

This paper presents a new method for representing DNA sequences. It permits the representation and investigation of patterns in sequences, visually revealing previously unknown structures. Based on a technique from chaotic dynamics, the method produces a picture of a gene sequence which displays both local and global patterns. The pictures have a complex structure which varies depending on the sequence. The method is termed Chaos Game Representation (CGR). CGR raises a new set of questions about the structure of DNA sequences, and is a new tool for investigating gene structure.