A knowledge base for predicting protein localization sites in eukaryotic cells

DeepAIP: Deep learning for anti-inflammatory peptide prediction using pre-trained protein language model features based on contextual self-attention network

Non-steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids, and other immunosuppressants are commonly used medications for treating inflammation. However, these drugs often come with numerous side effects. Therefore, finding more effective methods for inflammation treatment has become more necessary. The study of anti-inflammatory peptides can effectively address these issues. In this work, we propose a contextual self-attention deep learning model, coupled with features extracted from a pre-trained protein language model, to predict Anti-inflammatory Peptides (AIP). The contextual self-attention module can effectively enhance and learn the features extracted from the pre-trained protein language model, resulting in high accuracy to predict AIP. Additionally, we compared the performance of features extracted from popular pre-trained protein language models available in the market. Finally, Prot-T5 features demonstrated the best comprehensive performance as the input for our deep learning model named DeepAIP. Compared with existing methods on benchmark test dataset, DeepAIP gets higher Matthews Correlation Coefficient and Accuracy score than the second-best method by 16.35 % and 6.91 %, respectively. Performance comparison analysis was conducted using a dataset of 17 novel anti-inflammatory peptide sequences. DeepAIP demonstrates outstanding accuracy, correctly identifying all 17 peptide types as AIP and predicting values closer to the true ones. Data and code are available at https://github.com/YangQingGuoCCZU/DeepAIP.

Control of nuclear localization of the nucleocapsid protein of SARS-CoV-2

The nucleocapsid (N) protein of coronaviruses is a structural protein that binds viral RNA for assembly into the mature virion, a process that occurs in the cytoplasm. Several coronavirus N proteins also localize to the nucleus. Herein, we identify that two sequences (NLSs) are required for nuclear localization of the SARS-CoV-2 N protein. Deletion or mutation of these two sequences creates an N protein that does not localize to the nucleus in HEK293T cells. Overexpression of both wild-type and NLS-mutated N proteins dysregulate a largely overlapping set of mRNAs in HEK293T cells, suggesting that these N proteins do not have direct nuclear effects on transcription. Consistent with that hypothesis, both N proteins induce nuclear localization of NF-κB p65 and dysregulate a set of previously identified NF-κB-dependent genes. The effects of N on nuclear properties are proposed to alter host cell functions that contribute to viral pathogenesis or replication.

Translocation of Antimicrobial Peptides across Model Membranes: The Role of Peptide Chain Length

Cushioned lipid bilayers are structures consisting of a lipid bilayer supported on a solid substrate with an intervening layer of soft material. They offer possibilities for studying the behavior and interactions of biological membranes more accurately under physiological conditions. In this work, we continue our studies of cushion formation induced by histatin 5 (24Hst5), focusing on the effect of the length of the peptide chain. 24Hst5 is a short, positively charged, intrinsically disordered saliva peptide, and here, both a shorter (14Hst5) and a longer (48Hst5) peptide variant were evaluated. Experimental surface active techniques were combined with coarse-grained Monte Carlo simulations to obtain information about these peptides. Results show that at 10 mM NaCl, both the shorter and the longer peptide variants behave like 24Hst5 and a cushion below the bilayer is formed. At 150 mM NaCl, however, no interaction is observed for 24Hst5. On the contrary, a cushion is formed both in the case of 14Hst5 and 48Hst5, and in the latter, an additional thick, diffuse, and highly hydrated layer of peptide and lipid molecules is formed, on top of the bilayer. Similar trends were observed from the simulations, which allowed us to hypothesize that positively charged patches of the amino acids lysine and arginine in all three peptides are essential for them to interact with and translocate over the bilayer. We therefore hypothesize that electrostatic interactions are important for the interaction between the solid-supported lipid bilayers and the peptide depending on the linear charge density through the primary sequence and the positively charged patches in the sequence. The understanding of how, why, and when the cushion is formed opens up the possibility for this system to be used in the research and development of new drugs and pharmaceuticals.

Identification and Expression Analysis of UPS Gene Family in Potato

Ureide permeases (UPSs) mediate the transport of ureides, including allantoin and allantoate, which act as nitrogen-transporting compounds in plants and have recently been found to play a role in cellular signaling. To date, UPSs have not been reported in potato, and their identification is important for further function studies and for understanding molecular mechanisms of plant adverse responses. Based on potato genomic data, we identified 10 StUPS genes in potato (Solanum tuberosum L.). Then, we conducted a comprehensive study of the identified StUPS genes using bioinformatics methods. Genome phylogenetic and genomic localization analyses revealed that StUPSs can be classified into four categories, are highly homologous to Arabidopsis thaliana UPS members, and are distributed on three chromosomes. The six StUPS genes were investigated by RT-qPCR, and the findings indicated that all of these genes are involved in the response to several stresses, including low nitrogen, cold, ABA, salt, H2O2, and drought. This study establishes a strong theoretical framework for investigating the function of potato UPS genes, as well as the molecular mechanisms underlying the responses of these genes to various environmental stresses.

Altered socio-affective communication and amygdala development in mice with protocadherin10-deficient interneurons

Autism spectrum disorder (ASD) is a group of neurodevelopmental conditions associated with deficits in social interaction and communication, together with repetitive behaviours. The cell adhesion molecule protocadherin10 (PCDH10) is linked to ASD in humans. Pcdh10 is expressed in the nervous system during embryonic and early postnatal development and is important for neural circuit formation. In mice, strong expression of Pcdh10 in the ganglionic eminences and in the basolateral complex (BLC) of the amygdala was observed at mid and late embryonic stages, respectively. Both inhibitory and excitatory neurons expressed Pcdh10 in the BLC at perinatal stages and vocalization-related genes were enriched in Pcdh10-expressing neurons in adult mice. An epitope-tagged Pcdh10-HAV5 mouse line revealed endogenous interactions of PCDH10 with synaptic proteins in the young postnatal telencephalon. Nuanced socio-affective communication changes in call emission rates, acoustic features and call subtype clustering were primarily observed in heterozygous pups of a conditional knockout (cKO) with selective deletion of Pcdh10 in Gsh2-lineage interneurons. These changes were less prominent in heterozygous ubiquitous Pcdh10 KO pups, suggesting that altered anxiety levels associated with Gsh2-lineage interneuron functioning might drive the behavioural effects. Together, loss of Pcdh10 specifically in interneurons contributes to behavioural alterations in socio-affective communication with relevance to ASD.

Fast and exact fixed-radius neighbor search based on sorting

Fixed-radius near neighbor search is a fundamental data operation that retrieves all data points within a user-specified distance to a query point. There are efficient algorithms that can provide fast approximate query responses, but they often have a very compute-intensive indexing phase and require careful parameter tuning. Therefore, exact brute force and tree-based search methods are still widely used. Here we propose a new fixed-radius near neighbor search method, called SNN, that significantly improves over brute force and tree-based methods in terms of index and query time, provably returns exact results, and requires no parameter tuning. SNN exploits a sorting of the data points by their first principal component to prune the query search space. Further speedup is gained from an efficient implementation using high-level basic linear algebra subprograms (BLAS). We provide theoretical analysis of our method and demonstrate its practical performance when used stand-alone and when applied within the DBSCAN clustering algorithm.

Structural basis of nuclear transport for NEIL DNA glycosylases mediated by importin-alpha

NEIL glycosylases, including NEIL1, NEIL2, and NEIL3, play a crucial role in the base excision DNA repair pathway (BER). The classical importin pathway mediated by importin α/β and cargo proteins containing nuclear localization sequences (NLS) is the most common transport mechanism of DNA repair proteins to the nucleus. Previous studies have identified putative NLSs located at the C-terminus of NEIL3 and NEIL1. Crystallographic, bioinformatics, calorimetric (ITC), and fluorescence assays were used to investigate the interaction between NEIL1 and NEIL3 putative NLSs and importin-α (Impα). Our findings showed that NEIL3 contains a typical cNLS, with medium affinity for the major binding site of Impα. In contrast, crystallographic analysis of NEIL1 NLS revealed its binding to Impα, but with high B-factors and a lack of electron density at the linker region. ITC and fluorescence assays indicated no detectable affinity between NEIL1 NLS and Impα. These data suggest that NEIL1 NLS is a non-classical NLS with low affinity to Impα. Additionally, we compared the binding mode of NEIL3 and NEIL1 with Mus musculus Impα to human isoforms HsImpα1 and HsImpα3, which revealed interesting binding differences for HsImpα3 variant. NEIL3 is a classical medium affinity monopartite NLS, while NEIL1 is likely to be an unclassical low-affinity bipartite NLS. The base excision repair pathway is one of the primary systems involved in repairing DNA. Thus, understanding the mechanisms of nuclear transport of NEIL proteins is crucial for comprehending the role of these proteins in DNA repair and disease development.

A review from biological mapping to computation-based subcellular localization

Subcellular localization is crucial to the study of virus and diseases. Specifically, research on protein subcellular localization can help identify clues between virus and host cells that can aid in the design of targeted drugs. Research on RNA subcellular localization is significant for human diseases (such as Alzheimer's disease, colon cancer, etc.). To date, only reviews addressing subcellular localization of proteins have been published, which are outdated for reference, and reviews of RNA subcellular localization are not comprehensive. Therefore, we collated (the most up-to-date) literature on protein and RNA subcellular localization to help researchers understand changes in the field of protein and RNA subcellular localization. Extensive and complete methods for constructing subcellular localization models have also been summarized, which can help readers understand the changes in application of biotechnology and computer science in subcellular localization research and explore how to use biological data to construct improved subcellular localization models. This paper is the first review to cover both protein subcellular localization and RNA subcellular localization. We urge researchers from biology and computational biology to jointly pay attention to transformation patterns, interrelationships, differences, and causality of protein subcellular localization and RNA subcellular localization.

Epac as a tractable therapeutic target

In 1957, cyclic adenosine monophosphate (cAMP) was identified as the first secondary messenger, and the first signaling cascade discovered was the cAMP-protein kinase A (PKA) pathway. Since then, cAMP has received increasing attention given its multitude of actions. Not long ago, a new cAMP effector named exchange protein directly activated by cAMP (Epac) emerged as a critical mediator of cAMP's actions. Epac mediates a plethora of pathophysiologic processes and contributes to the pathogenesis of several diseases such as cancer, cardiovascular disease, diabetes, lung fibrosis, neurological disorders, and others. These findings strongly underscore the potential of Epac as a tractable therapeutic target. In this context, Epac modulators seem to possess unique characteristics and advantages and hold the promise of providing more efficacious treatments for a wide array of diseases. This paper provides an in-depth dissection and analysis of Epac structure, distribution, subcellular compartmentalization, and signaling mechanisms. We elaborate on how these characteristics can be utilized to design specific, efficient, and safe Epac agonists and antagonists that can be incorporated into future pharmacotherapeutics. In addition, we provide a detailed portfolio for specific Epac modulators highlighting their discovery, advantages, potential concerns, and utilization in the context of clinical disease entities.

Saliency-Based Multilabel Linear Discriminant Analysis

Linear discriminant analysis (LDA) is a classical statistical machine-learning method, which aims to find a linear data transformation increasing class discrimination in an optimal discriminant subspace. Traditional LDA sets assumptions related to the Gaussian class distributions and single-label data annotations. In this article, we propose a new variant of LDA to be used in multilabel classification tasks for dimensionality reduction on original data to enhance the subsequent performance of any multilabel classifier. A probabilistic class saliency estimation approach is introduced for computing saliency-based weights for all instances. We use the weights to redefine the between-class and within-class scatter matrices needed for calculating the projection matrix. We formulate six different variants of the proposed saliency-based multilabel LDA (SMLDA) based on different prior information on the importance of each instance for their class(es) extracted from labels and features. Our experiments show that the proposed SMLDA leads to performance improvements in various multilabel classification problems compared to several competing dimensionality reduction methods.

Plasmodium berghei Brca2 is required for normal development and differentiation in mice and mosquitoes

BACKGROUND

Malaria is a major global parasitic disease caused by species of the genus Plasmodium. Zygotes of Plasmodium spp. undergo meiosis and develop into tetraploid ookinetes, which differentiate into oocysts that undergo sporogony. Homologous recombination (HR) occurs during meiosis and introduces genetic variation. However, the mechanisms of HR in Plasmodium are unclear. In humans, the recombinases DNA repair protein Rad51 homolog 1 (Rad51) and DNA meiotic recombinase 1 (Dmc1) are required for HR and are regulated by breast cancer susceptibility protein 2 (BRCA2). Most eukaryotes harbor BRCA2 homologs. Nevertheless, these have not been reported for Plasmodium.

METHODS

A Brca2 candidate was salvaged from a database to identify Brca2 homologs in Plasmodium. To confirm that the candidate protein was Brca2, interaction activity between Plasmodium berghei (Pb) Brca2 (PbBrca2) and Rad51 (PbRad51) was investigated using a mammalian two-hybrid assay. To elucidate the functions of PbBrca2, PbBrca2 was knocked out and parasite proliferation and differentiation were assessed in mice and mosquitoes. Transmission electron microscopy was used to identify sporogony.

RESULTS

The candidate protein was conserved among Plasmodium species, and it was indicated that it harbors critical BRCA2 domains including BRC repeats, tower, and oligonucleotide/oligosaccharide-binding-fold domains. The P. berghei BRC repeats interacted with PbRad51. Hence, the candidate was considered a Brca2 homolog. PbBrca2 knockout parasites were associated with reduced parasitemia with increased ring stage and decreased trophozoite stage counts, gametocytemia, female gametocyte ratio, oocyst number, and ookinete development in both mice and mosquitoes. Nevertheless, the morphology of the blood stages in mice and the ookinete stage was comparable to those of the wild type parasites. Transmission electron microscopy results showed that sporogony never progressed in Brca2-knockout parasites.

CONCLUSIONS

Brca2 is implicated in nearly all Plasmodium life cycle stages, and especially in sporogony. PbBrca2 contributes to HR during meiosis.

Optimization of production and characterization of a recombinant soluble human Cripto-1 protein inhibiting self-renewal of cancer stem cells

Human Cripto-1 is a member of the epidermal growth factor (EGF)-Cripto-FRL-1-Cryptic (CFC) family family and performs critical roles in cancer and various pathological and developmental processes. Recently we demonstrated that a soluble form of Cripto-1 suppresses the self-renewal and enhances the differentiation of cancer stem cells (CSCs). A functional form of soluble Cripto-1 was found to be difficult to obtain because of the 12 cysteine residues in the protein which impairs the folding process. Here, we optimized the protocol for a T7 expression system, purification from inclusion bodies under denatured conditions refolding of a His-tagged Cripto-1 protein. A concentrations of 0.2-0.4 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) at 37°C was found to be the optimal concentration for Cripto-1 expression while imidazole at 0.5 M was the optimum concentration to elute the Cripto-1 protein from a Ni-column in the smallest volume. Cation exchange column chromatography of the Cripto-1 protein in the presence of 8 M urea exhibited sufficient elution profile at pH 5, which was more efficient at recovery. The recovery of the protein reached to more than 26.6% after refolding with arginine. The purified Cripto-1 exhibited high affinity to the anti-ALK-4 antibody and suppressed sphere forming ability of CSCs at high dose and induced cell differentiation.

Genomic and Experimental Analysis of the Insecticidal Factors Secreted by the Entomopathogenic Fungus Beauveria pseudobassiana RGM 2184

The entomopathogenic fungus Beauveria pseudobassiana strain RGM 2184 can reach a maximum efficacy of 80% against the quarantine pest Lobesia botrana in field assays. In this study, the RGM 2184 genome was sequenced, and genome mining analyses were performed to predict the factors involved in its insecticidal activity. Additionally, the metabolic profiling of the RMG 2184 culture's supernatants was analyzed by mass spectrometry, and the insecticidal activity from one of these extracts was evaluated in Galleria mellonella larvae. The genome analysis resulted in 114 genes encoding for extracellular enzymes, four biosynthetic gene clusters reported as producers of insecticidal and bactericidal factors (oosporein, beauvericin, desmethylbassianin, and beauveriolide), 20 toxins, and at least 40 undescribed potential biocontrol factors (polyketides and nonribosomal peptides). Comparative genomic analysis revealed that 65-95% of these genes are Beauveria genus-specific. Metabolic profiling of supernatant extracts from RGM 2184 cultures exhibited secondary metabolites such as beauveriolide, oosporein, inflatin C, and bassiatin. However, a number of detected metabolites still remain undescribed. The metabolite extract caused 79% mortality of Galleria mellonella larvae at 28 days. The results of this research lay the groundwork for the study of new insecticidal molecules.

Minimum adjusted Rand index for two clusterings of a given size

The adjusted Rand index (ARI) is commonly used in cluster analysis to measure the degree of agreement between two data partitions. Since its introduction, exploring the situations of extreme agreement and disagreement under different circumstances has been a subject of interest, in order to achieve a better understanding of this index. Here, an explicit formula for the lowest possible value of the ARI for two clusterings of given sizes is shown, and moreover a specific pair of clusterings achieving such a bound is provided.

Karyopherin-mediated nucleocytoplasmic transport

Efficient and regulated nucleocytoplasmic trafficking of macromolecules to the correct subcellular compartment is critical for proper functions of the eukaryotic cell. The majority of the macromolecular traffic across the nuclear pores is mediated by the Karyopherin-β (or Kap) family of nuclear transport receptors. Work over more than two decades has shed considerable light on how the different Kap family members bring their respective cargoes into the nucleus or the cytoplasm in efficient and highly regulated manners. In this Review, we overview the main features and established functions of Kap family members, describe how Kaps recognize their cargoes and discuss the different ways in which these Kap-cargo interactions can be regulated, highlighting new findings and open questions. We also describe current knowledge of the import and export of the components of three large gene expression machines - the core replisome, RNA polymerase II and the ribosome - pointing out the questions that persist about how such large macromolecular complexes are trafficked to serve their function in a designated subcellular location.

Genetic structure and evolutionary diversity of mating-type (MAT) loci in Hypsizygus marmoreus

The mating compatibility in fungi is generally governed by genes located within a single or two unlinked mating type (MAT) loci. Hypsizygus marmoreus is an edible mushroom in the order Agaricales with a tetrapolar system, which contains two unlinked MAT loci-homeodomain (HD) transcription factor genes and pheromone/pheromone receptor genes (P/R). In this study, we analyzed the genetic structure and diversity of MAT loci in tetrapolar system of H. marmoreus through sequencing of 54 heterokaryon and 8 homokaryon strains. Although within the HD loci, the gene order was conserved, the gene contents were variable, and the HD loci haplotypes were further classified into four types. By analyzing the structure, phylogeny, and the HD transmissibility based on the progeny of these four HD mating-type loci types, we found that they were heritable and tightly linked at the HD loci. The P/R loci genes were found to comprise three pheromone receptors, three pheromones, and two pheromone receptor-like genes. Intra- and inter-specific phylogenetic analyses of pheromone receptors revealed that the STE3 genes were divided into three groups, and we thus theorize that they diverged before speciation. Comparative analysis of the MAT regions among 73 Basidiomycete species indicated that the diversity of HD and P/R loci in Agaricales and Boletales may contribute to mating compatibility. The number of HD genes were not correlated with the tetrapolar or bipolar systems. In H. marmoreus, the expression levels of these genes at HD and P/R loci of compatible strains were found higher than in those of homonuclear/homokaryotic strains, indicating that these mating genes acted as switches for mating processes. Further collinear analysis of HD loci in interspecific species found that HD loci contains conserved recombination hotspots showing major rearrangements in Coprinopsis cinerea and Schizophyllum commune, suggesting different mechanisms for evolution of physically linked MAT loci in these groups. It seems likely that gene rearrangements are common in Agaricales fungi around HD loci. Together, our study provides insights into the genomic basis of mating compatibility in H. marmoreus.

Role of two aminoacyl-tRNA synthetase associated proteins (Endothelial Monocyte Activating Polypeptides 1 and 2) of Leishmania donovani in chemotaxis of human monocytes

Visceral leishmaniasis is caused by the protozoan parasite Leishmania donovani. It is a fatal form of leishmaniasis prevalent in Indian subcontinent. Since there are no human licensed vaccines available for leishmaniasis, chemotherapeutic drugs remain the only means for combating parasitic infections. We have earlier identified a total of 26 amino-acyl tRNA synthetases (aaRS) along with five stand-alone editing domains and two aaRS-associated proteins in Leishmania donovani. In addition to their canonical role of tRNA aminoacylation, aaRS have been involved in novel functions by acquiring novel domains during evolution. The aaRS-associated proteins have been reported to be analogous to a human cytokine, EMAP II, as they possess a modified version of the heptapeptide motif responsible for the cytokine activity. In this manuscript, we report the characterization of two L. donovani aminoacyl-tRNA synthetase associated proteins which showed a human chemokine like activity. Both the proteins, L. donovani EMAP-1 and EMAP-2, possess a modified form of the heptapeptide motif, which is responsible for cytokine activity in human EMAP-2. LdEMAP-1 and LdEMAP-2 were cloned, expressed, and purified. Both LdEMAP-1 and LdEMAP-2 proteins in the promastigote stage were found to be localized in cytoplasm as confirmed by immunofluorescence. In case of L. donovani infected human THP-1 derived macrophages, secretion of LdEMAP-1 and LdEMAP-2 proteins in the cytosol of the macrophages was observed. The role of LdEMAP-1 and LdEMAP-2 in the aminoacylation of rLdTyrRS was also tested and LdEMAP-2 but not LdEMAP-1 increased the rate of aminoacylation of tyrosyl tRNA synthetase (rLdTyrRS). L. donovani EMAP-1 and EMAP-2 proteins managed to exhibit the capability of attracting human origin cells as determined by chemotaxis assay, and also were able to induce the secretion of cytokines from macrophages like their human counterpart (EMAP II). Our working hypothesis is that both of these proteins might be involved in helping the parasite to establish the infection within the host.

Summarizing Finite Mixture Model with Overlapping Quantification

Finite mixture models are widely used for modeling and clustering data. When they are used for clustering, they are often interpreted by regarding each component as one cluster. However, this assumption may be invalid when the components overlap. It leads to the issue of analyzing such overlaps to correctly understand the models. The primary purpose of this paper is to establish a theoretical framework for interpreting the overlapping mixture models by estimating how they overlap, using measures of information such as entropy and mutual information. This is achieved by merging components to regard multiple components as one cluster and summarizing the merging results. First, we propose three conditions that any merging criterion should satisfy. Then, we investigate whether several existing merging criteria satisfy the conditions and modify them to fulfill more conditions. Second, we propose a novel concept named clustering summarization to evaluate the merging results. In it, we can quantify how overlapped and biased the clusters are, using mutual information-based criteria. Using artificial and real datasets, we empirically demonstrate that our methods of modifying criteria and summarizing results are effective for understanding the cluster structures. We therefore give a new view of interpretability/explainability for model-based clustering.

Identifying carbohydrate-active enzymes of Cutaneotrichosporon oleaginosus using systems biology

Background

The oleaginous yeast Cutaneotrichosporon oleaginosus represents one of the most promising microbial platforms for resource-efficient and scalable lipid production, with the capacity to accept a wide range of carbohydrates encapsulated in complex biomass waste or lignocellulosic hydrolysates. Currently, data related to molecular aspects of the metabolic utilisation of oligomeric carbohydrates are sparse. In addition, comprehensive proteomic information for C. oleaginosus focusing on carbohydrate metabolism is not available.

Results

In this study, we conducted a systematic analysis of carbohydrate intake and utilisation by C. oleaginosus and investigated the influence of different di- and trisaccharide as carbon sources. Changes in the cellular growth and morphology could be observed, depending on the selected carbon source. The greatest changes in morphology were observed in media containing trehalose. A comprehensive proteomic analysis of secreted, cell wall-associated, and cytoplasmatic proteins was performed, which highlighted differences in the composition and quantity of secreted proteins, when grown on different disaccharides. Based on the proteomic data, we performed a relative quantitative analysis of the identified proteins (using glucose as the reference carbon source) and observed carbohydrate-specific protein distributions. When using cellobiose or lactose as the carbon source, we detected three- and five-fold higher diversity in terms of the respective hydrolases released. Furthermore, the analysis of the secreted enzymes enabled identification of the motif with the consensus sequence LALL[LA]L[LA][LA]AAAAAAA as a potential signal peptide.

Conclusions

Relative quantification of spectral intensities from crude proteomic datasets enabled the identification of new enzymes and provided new insights into protein secretion, as well as the molecular mechanisms of carbo-hydrolases involved in the cleavage of the selected carbon oligomers. These insights can help unlock new substrate sources for C. oleaginosus, such as low-cost by-products containing difficult to utilize carbohydrates. In addition, information regarding the carbo-hydrolytic potential of C. oleaginosus facilitates a more precise engineering approach when using targeted genetic approaches. This information could be used to find new and more cost-effective carbon sources for microbial lipid production by the oleaginous yeast C. oleaginosus.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12934-021-01692-2.

Predicting subcellular location of protein with evolution information and sequence-based deep learning

BACKGROUND

Protein subcellular localization prediction plays an important role in biology research. Since traditional methods are laborious and time-consuming, many machine learning-based prediction methods have been proposed. However, most of the proposed methods ignore the evolution information of proteins. In order to improve the prediction accuracy, we present a deep learning-based method to predict protein subcellular locations.

RESULTS

Our method utilizes not only amino acid compositions sequence but also evolution matrices of proteins. Our method uses a bidirectional long short-term memory network that processes the entire protein sequence and a convolutional neural network that extracts features from protein sequences. The position specific scoring matrix is used as a supplement to protein sequences. Our method was trained and tested on two benchmark datasets. The experiment results show that our method yields accurate results on the two datasets with an average precision of 0.7901, ranking loss of 0.0758 and coverage of 1.2848.

CONCLUSION

The experiment results show that our method outperforms five methods currently available. According to those experiments, we can see that our method is an acceptable alternative to predict protein subcellular location.

ImbTreeEntropy and ImbTreeAUC: Novel R Packages for Decision Tree Learning on the Imbalanced Datasets

This paper presents two R packages ImbTreeEntropy and ImbTreeAUC to handle imbalanced data problems. ImbTreeEntropy functionality includes application of a generalized entropy functions, such as Rényi, Tsallis, Sharma–Mittal, Sharma–Taneja and Kapur, to measure impurity of a node. ImbTreeAUC provides non-standard measures to choose an optimal split point for an attribute (as well the optimal attribute for splitting) by employing local, semi-global and global AUC (Area Under the ROC curve) measures. Both packages are applicable for binary and multiclass problems and they support cost-sensitive learning, by defining a misclassification cost matrix, and weighted-sensitive learning. The packages accept all types of attributes, including continuous, ordered and nominal, where the latter type is simplified for multiclass problems to reduce the computational overheads. Both applications enable optimization of the thresholds where posterior probabilities determine final class labels in a way that misclassification costs are minimized. Model overfitting can be managed either during the growing phase or at the end using post-pruning. The packages are mainly implemented in R, however some computationally demanding functions are written in plain C++. In order to speed up learning time, parallel processing is supported as well.

A Novel Multiprotein Bridging Factor 1-Like Protein Induces Cyst Wall Protein Gene Expression and Cyst Differentiation in Giardia lamblia

The capacity to synthesize a protective cyst wall is critical for infectivity of Giardia lamblia. It is of interest to know the mechanism of coordinated synthesis of three cyst wall proteins (CWPs) during encystation, a differentiation process. Multiprotein bridging factor 1 (MBF1) gene family is a group of transcription coactivators that bridge various transcription factors. They are involved in cell growth and differentiation in yeast and animals, or in stress response in fungi and plants. We asked whether Giardia has MBF1-like genes and whether their products influence gene expression. BLAST searches of the Giardia genome database identified one gene encoding a putative MBF1 protein with a helix-turn-helix domain. We found that it can specifically bind to the AT-rich initiator promoters of the encystation-induced cwp1-3 and myb2 genes. MBF1 localized to cell nuclei and cytoplasm with higher expression during encystation. In addition, overexpression of MBF1 induced cwp1-3 and myb2 gene expression and cyst generation. Mutation of the helixes in the helix-turn-helix domain reduced cwp1-3 and myb2 gene expression and cyst generation. Chromatin immunoprecipitation assays confirmed the binding of MBF1 to the promoters with its binding sites in vivo. We also found that MBF1 can interact with E2F1, Pax2, WRKY, and Myb2 transcription factors that coordinately up-regulate the cwp genes during encystation. Using a CRISPR/Cas9 system for targeted disruption of mbf1 gene, we found a downregulation of cwp1-3 and myb2 genes and decrease of cyst generation. Our results suggest that MBF1 is functionally conserved and positively regulates Giardia cyst differentiation.

Mitochondrial protein phosphorylation in yeast revisited

Protein phosphorylation is one of the best-known post-translational modifications occurring in all domains of life. In eukaryotes, protein phosphorylation affects all cellular compartments including mitochondria. High-throughput techniques of mass spectrometry combined with cell fractionation and biochemical methods yielded thousands of phospho-sites on hundreds of mitochondrial proteins. We have compiled the information on mitochondrial protein kinases and phosphatases and their substrates in Saccharomyces cerevisiae and provide the current state-of-the-art overview of mitochondrial protein phosphorylation in this model eukaryote. Using several examples, we describe emerging features of the yeast mitochondrial phosphoproteome and present challenges lying ahead in this exciting field.

Identification and characterization of CTP:phosphocholine cytidylyltransferase CpCCT1 in the resurrection plant Craterostigma plantagineum

Phosphatidylcholine is a major phospholipid which is shown to be involved in stress adaptation. Phosphatidylcholine increased during dehydration in Craterostigma plantagineum, and therefore we characterized CTP:phosphocholine cytidylyltransferase (CpCCT1), a key regulatory enzyme for phosphatidylcholine synthesis in plants. The CpCCT1 gene from the resurrection plant C. plantagineum was cloned and the amino acid sequence was compared with homologs from other species including yeast and rat. CCT proteins have conserved catalytic and membrane-binding domains while the N-terminal and C-terminal domains have diverged. The tissue specific expression analysis indicated that CpCCT1 is expressed in all tested tissues and it is induced by dehydration and in response to 0.5 M NaCl solutions. In plants exposed to low temperature in the dark, the CpCCT1 transcript increased after 4 h at 4 °C. CpCCT1 expression also increased during mannitol and sorbitol treatments in a concentration dependent manner. Phytohormones such as abscisic acid and indole-3-acetic acid also trigged transcript accumulation. Comparisons of transcript and protein accumulations for different treatments (except for dehydration) suggest transcriptional and translational control mechanisms. Analysis of promoter activity and polysome occupancy suggest that CpCCT1 gene expression is mainly under translational regulation during dehydration.

Protein Subcellular Localization Prediction

The elucidation of the subcellular localization of proteins is very important in order to deeply understand their functions. In fact, proteins activities are strictly correlated to the cellular compartment and microenvironment in which they are present.In recent years, several effective and reliable proteomics techniques and computational methods have been developed and implemented in order to identify the proteins subcellular localization. This process is often time-consuming and expensive, but the recent technological and bioinformatics progress allowed the development of more accurate and simple workflows to determine the localization, interactions, and functions of proteins.In the following chapter, a brief introduction on the importance of knowing subcellular localization of proteins will be presented. Then, sample preparation protocols, proteomic methods, data analysis strategies, and software for the prediction of proteins localization will be presented and discussed. Finally, the more recent and advanced spatial proteomics techniques will be shown.

Improvement of fatty acid productivity of thraustochytrid, Aurantiochytrium sp. by genome editing

Thraustochytrid strains belonging to the genus Aurantiochytrium accumulate significant amounts of lipids including polyunsaturated fatty acids and carotenoids and, therefore, are expected to be used for industrial production of various valuable materials. Although various efforts such as chemical mutagenesis and homologous gene recombination have been made to improve lipid productivity of Aurantiochytrium species, low specificity and efficiency in the conventional methods hinder the research progress. Here, we attempted to apply a genome editing technology, the CRISPR-Cas9 system as an alternative molecular breeding technique for Aurantiochytrium species to accelerate the metabolic engineering. The efficiency of specific gene knock-in by the homologous recombination increased more than 10-folds by combining the CRISPR-Cas9 system. As a result of disrupting the genes associated with β-oxidation of fatty acids by the improved method, the genome edited strains with higher fatty acid productivity were isolated, demonstrating for the first time that the CRISPR-Cas9 system was effective for molecular breeding of the strains in the genus Aurantiochytrium to improve lipid productivity.

Unravelling the Network of Nuclear Matrix Metalloproteinases for Targeted Drug Design

Matrix metalloproteinases (MMPs) are zinc-dependent endopeptidases that are responsible for the degradation of a wide range of extracellular matrix proteins, which are involved in many cellular processes to ensure the normal development of tissues and organs. Overexpression of MMPs has been observed to facilitate cellular growth, migration, and metastasis of tumor cells during cancer progression. A growing number of these proteins are being found to exist in the nuclei of both healthy and tumor cells, thus highlighting their localization as having a genuine purpose in cellular homeostasis. The mechanism underlying nuclear transport and the effects of MMP nuclear translocation have not yet been fully elucidated. To date, nuclear MMPs appear to have a unique impact on cellular apoptosis and gene regulation, which can have effects on immune response and tumor progression, and thus present themselves as potential therapeutic targets in certain types of cancer or disease. Herein, we highlight and evaluate what progress has been made in this area of research, which clearly has some value as a specific and unique way of targeting the activity of nuclear matrix metalloproteinases within various cell types.

Comprehensive Structural and Molecular Comparison of Spike Proteins of SARS-CoV-2, SARS-CoV and MERS-CoV, and Their Interactions with ACE2

The severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has recently emerged in China and caused a disease called coronavirus disease 2019 (COVID-19). The virus quickly spread around the world, causing a sustained global outbreak. Although SARS-CoV-2, and other coronaviruses, SARS-CoV and Middle East respiratory syndrome CoV (MERS-CoV) are highly similar genetically and at the protein production level, there are significant differences between them. Research has shown that the structural spike (S) protein plays an important role in the evolution and transmission of SARS-CoV-2. So far, studies have shown that various genes encoding primarily for elements of S protein undergo frequent mutation. We have performed an in-depth review of the literature covering the structural and mutational aspects of S protein in the context of SARS-CoV-2, and compared them with those of SARS-CoV and MERS-CoV. Our analytical approach consisted in an initial genome and transcriptome analysis, followed by primary, secondary and tertiary protein structure analysis. Additionally, we investigated the potential effects of these differences on the S protein binding and interactions to angiotensin-converting enzyme 2 (ACE2), and we established, after extensive analysis of previous research articles, that SARS-CoV-2 and SARS-CoV use different ends/regions in S protein receptor-binding motif (RBM) and different types of interactions for their chief binding with ACE2. These differences may have significant implications on pathogenesis, entry and ability to infect intermediate hosts for these coronaviruses. This review comprehensively addresses in detail the variations in S protein, its receptor-binding characteristics and detailed structural interactions, the process of cleavage involved in priming, as well as other differences between coronaviruses.

Tools for the Recognition of Sorting Signals and the Prediction of Subcellular Localization of Proteins From Their Amino Acid Sequences

At the time of translation, nascent proteins are thought to be sorted into their final subcellular localization sites, based on the part of their amino acid sequences (i.e., sorting or targeting signals). Thus, it is interesting to computationally recognize these signals from the amino acid sequences of any given proteins and to predict their final subcellular localization with such information, supplemented with additional information (e.g., k-mer frequency). This field has a long history and many prediction tools have been released. Even in this era of proteomic atlas at the single-cell level, researchers continue to develop new algorithms, aiming at accessing the impact of disease-causing mutations/cell type-specific alternative splicing, for example. In this article, we overview the entire field and discuss its future direction.

GP4: an integrated Gram-Positive Protein Prediction Pipeline for subcellular localization mimicking bacterial sorting

Subcellular localization is a critical aspect of protein function and the potential application of proteins either as drugs or drug targets, or in industrial and domestic applications. However, the experimental determination of protein localization is time consuming and expensive. Therefore, various localization predictors have been developed for particular groups of species. Intriguingly, despite their major representation amongst biotechnological cell factories and pathogens, a meta-predictor based on sorting signals and specific for Gram-positive bacteria was still lacking. Here we present GP⁴, a protein subcellular localization meta-predictor mainly for Firmicutes, but also Actinobacteria, based on the combination of multiple tools, each specific for different sorting signals and compartments. Novelty elements include improved cell-wall protein prediction, including differentiation of the type of interaction, prediction of non-canonical secretion pathway target proteins, separate prediction of lipoproteins and better user experience in terms of parsability and interpretability of the results. GP⁴ aims at mimicking protein sorting as it would happen in a bacterial cell. As GP⁴ is not homology based, it has a broad applicability and does not depend on annotated databases with homologous proteins. Non-canonical usage may include little studied or novel species, synthetic and engineered organisms, and even re-use of the prediction data to develop custom prediction algorithms. Our benchmark analysis highlights the improved performance of GP⁴ compared to other widely used subcellular protein localization predictors. A webserver running GP⁴ is available at http://gp4.hpc.rug.nl/

Use of Chou's 5-steps rule to predict the subcellular localization of gram-negative and gram-positive bacterial proteins by multi-label learning based on gene ontology annotation and profile alignment

To date, many proteins generated by large-scale genome sequencing projects are still uncharacterized and subject to intensive investigations by both experimental and computational means. Knowledge of protein subcellular localization (SCL) is of key importance for protein function elucidation. However, it remains a challenging task, especially for multiple sites proteins known to shuttle between cell compartments to perform their proper biological functions and proteins which do not have significant homology to proteins of known subcellular locations. Due to their low-cost and reasonable accuracy, machine learning-based methods have gained much attention in this context with the availability of a plethora of biological databases and annotated proteins for analysis and benchmarking. Various predictive models have been proposed to tackle the SCL problem, using different protein sequence features pertaining to the subcellular localization, however, the overwhelming majority of them focuses on single localization and cover very limited cellular locations. The prediction was basically established on sorting signals, amino acids compositions, and homology. To improve the prediction quality, focus is actually on knowledge information extracted from annotation databases, such as protein-protein interactions and Gene Ontology (GO) functional domains annotation which has been recently a widely adopted and essential information for learning systems. To deal with such problem, in the present study, we considered SCL prediction task as a multi-label learning problem and tried to label both single site and multiple sites unannotated bacterial protein sequences by mining proteins homology relationships using both GO terms of protein homologs and PSI-BLAST profiles. The experiments using 5-fold cross-validation tests on the benchmark datasets showed a significant improvement on the results obtained by the proposed consensus multi-label prediction model which discriminates six compartments for Gram-negative and five compartments for Gram-positive bacterial proteins.

Error Detection in a Large-Scale Lexical Taxonomy

Knowledge base (KB) is an important aspect in artificial intelligence. One significant challenge faced by KB construction is that it contains many noises, which prevent its effective usage. Even though some KB cleansing algorithms have been proposed, they focus on the structure of the knowledge graph and neglect the relation between the concepts, which could be helpful to discover wrong relations in KB. Motived by this, we measure the relation of two concepts by the distance between their corresponding instances and detect errors within the intersection of the conflicting concept sets. For efficient and effective knowledge base cleansing, we first apply a distance-based model to determine the conflicting concept sets using two different methods. Then, we propose and analyze several algorithms on how to detect and repair the errors based on our model, where we use a hash method for an efficient way to calculate distance. Experimental results demonstrate that the proposed approaches could cleanse the knowledge bases efficiently and effectively.

Discovery of novel [FeFe]-hydrogenases for biocatalytic H2-production

A semi-synthetic screening method for mining the biodiversity of [FeFe]-hydrogenases, expanding the toolbox for biocatalytic H₂-gas production.

A new screening method for [FeFe]-hydrogenases is described, circumventing the need for specialized expression conditions as well as protein purification for initial characterization. [FeFe]-hydrogenases catalyze the formation and oxidation of molecular hydrogen at rates exceeding 10³ s^–1, making them highly promising for biotechnological applications. However, the discovery of novel [FeFe]-hydrogenases is slow due to their oxygen sensitivity and dependency on a structurally unique cofactor, complicating protein expression and purification. Consequently, only a very limited number have been characterized, hampering their implementation. With the purpose of increasing the throughput of [FeFe]-hydrogenase discovery, we have developed a screening method that allows for rapid identification of novel [FeFe]-hydrogenases as well as their characterization with regards to activity (activity assays and protein film electrochemistry) and spectroscopic properties (electron paramagnetic resonance and Fourier transform infrared spectroscopy). The method is based on in vivo artificial maturation of [FeFe]-hydrogenases in Escherichia coli and all procedures are performed on either whole cells or non-purified cell lysates, thereby circumventing extensive protein purification. The screening was applied on eight putative [FeFe]-hydrogenases originating from different structural sub-classes and resulted in the discovery of two new active [FeFe]-hydrogenases. The [FeFe]-hydrogenase from Solobacterium moorei shows high H₂-gas production activity, while the enzyme from Thermoanaerobacter mathranii represents a hitherto uncharacterized [FeFe]-hydrogenase sub-class. This latter enzyme is a putative sensory hydrogenase and our in vivo spectroscopy study reveals distinct differences compared to the well established H₂ producing HydA1 hydrogenase from Chlamydomonas reinhardtii.

Identification of peptide domains involved in the subcellular localization of the feline coronavirus 3b protein

Feline coronavirus (FCoV) has been identified as the aetiological agent of feline infectious peritonitis (FIP), a highly fatal systemic disease in cats. FCoV open reading frame 3 (ORF3) encodes accessory proteins 3a, 3b and 3 c. The FCoV 3b accessory protein consists of 72 amino acid residues and localizes to nucleoli and mitochondria. The present work focused on peptide domains within FCoV 3b that drive its intracellular trafficking. Transfection of different cell types with FCoV 3b fused to enhanced green fluorescent protein (EGFP) or 3×FLAG confirmed localization of FCoV 3b in the mitochondria and nucleoli. Using serial truncated mutants, we showed that nucleolar accumulation is controlled by a joint nucleolar and nuclear localization signal (NoLS/NLS) in which the identified overlapping pat4 motifs (residues 53–57) play a critical role. Mutational analysis also revealed that mitochondrial translocation is mediated by N-terminal residues 10–35, in which a Tom20 recognition motif (residues 13–17) and two other overlapping hexamers (residues 24–30) associated with mitochondrial targeting were identified. In addition, a second Tom20 recognition motif was identified further downstream (residues 61–65), although the mitochondrial translocation evoked by these residues seemed less efficient as a diffuse cytoplasmic distribution was also observed. Assessing the spatiotemporal distribution of FCoV 3b did not provide convincing evidence of dynamic shuttling behaviour between the nucleoli and the mitochondria.

Plasma membrane damage repair is mediated by an acid sphingomyelinase in Entamoeba histolytica

Entamoeba histolytica is a pathogen that during its infective process confronts the host defenses, which damages the amoebic plasma membrane (PM), resulting in the loss of viability. However, it is unknown whether amoebic trophozoites are able to repair their PM when it is damaged. Acid sphingomyelinases (aSMases) have been reported in mammalian cells to promote endocytosis and removal of PM lesions. In this work, six predicted amoebic genes encoding for aSMases were found to be transcribed in the HM1:IMSS strain, finding that the EhaSM6 gene is the most transcribed in basal growth conditions and rendered a functional protein. The secreted aSMase activity detected was stimulated by Mg⁺² and inhibited by Co⁺². Trophozoites that overexpress the EhaSM6 gene (HM1-SM6HA) exhibit an increase of 2-fold in the secreted aSMase activity. This transfectant trophozoites exposed to pore-forming molecules (SLO, Magainin, β-Defensin 2 and human complement) exhibited an increase from 6 to 25-fold in the secreted aSMase activity which correlated with higher amoebic viability in a Ca⁺² dependent process. However, other agents that affect the PM such as hydrogen peroxide also induced an increase of secreted aSMase, but to a lesser extent. The aSMase6 enzyme is N- and C-terminal processed. Confocal and transmission electron microscopy showed that trophozoites treated with SLO presented a migration of lysosomes containing the aSMase towards the PM, inducing the formation of membrane patches and endosomes in the control strain. These cellular structures were increased in the overexpressing strain, indicating the involvement of the aSMase6 in the PM injury repair. The pore-forming molecules induced an increase in the expression of EhaSM1, 2, 5 and 6 genes, meanwhile, hydrogen peroxide induced an increase in all of them. In all the conditions evaluated, the EhaSM6 gene exhibited the highest levels of induction. Overall, these novel findings show that the aSMase6 enzyme from E. histolytica promotes the repair of the PM damaged with pore-forming molecules to prevent losing cell integrity. This novel system could act when encountered with the lytic defense systems of the host.

The host-amoeba relationship is based on a series of interplays between host defense mechanisms and parasite survival strategies. While host cells elaborate diverse mechanisms for pathogen elimination, Entamoeba histolytica trophozoites have also developed complex strategies to counteract host immune response and facilitate its own survival while confronting host defenses. E. histolytica exposed to pore-forming proteins such as β-Defensin 2, human complement and Streptolysin O (SLO), increases the activity of secreted aSMase, which is related to greater amoebic viability. Other agents that affect plasma membrane (PM) may also increase secreted aSMase but to a lesser extent. SLO form pores in the PM of E. histolytica trophozoites that initiates the uncontrolled entry of Ca²⁺, recognized as the primary trigger for cell responses which favors the migration of the lysosomes to the periphery of the cell, fuses with the PM and release their content, including aSMase to the external side of the cell. The secreted aSMase favoring the internalization of the lesion for its degradation in phagolysosomes. During the early stages of PM damage, the pores are rapidly blocked by patch-like structures that prevent the lysis of the trophozoite and immediately begin internalizing the lesion. The aSMase6 overexpression favors the repair of the lesion and the survival of E. histolytica trophozoites. Pore-forming proteins induced an increase in the expression of EhaSM1, 2, 5 and 6 genes, meanwhile oxidative stress induced an increase in all of them. Here we report, for the first time, that E. histolytica possess a mechanism for PM damage repair mediated by aSMase similar to the system described in mammalian cells.

The dual function gene RAD23 contributes to Cryptococcus neoformans virulence independently of its role in nucleotide excision DNA repair

Genes involved in the repair of DNA damage are emerging as playing important roles during the disease processes caused by pathogenic fungi. However, there are potentially hundreds of genes involved in DNA repair in a fungus and some of those genes can play additional roles within the cell. One such gene is RAD23, required for virulence of the human pathogenic fungus Cryptococcus neoformans, that encodes a protein involved in the nucleotide excision repair (NER) pathway. However, Rad23 is a dual function protein, with a role in either repair of damaged DNA or protein turn over by directing proteins to the proteasome. Here, these two functions of Rad23 were tested by the creation of a series of domain deletion alleles of RAD23 and the assessment of the strains for DNA repair, proteasome functions, and virulence properties. Deletion of the different domains was able to uncouple the two functions of Rad23, and the phenotypes of strains carrying such forms indicated that the role of RAD23 in virulence is due to its function in proteasomal-mediated protein degradation rather than NER.

Benchmarking subcellular localization and variant tolerance predictors on membrane proteins

Background

Membrane proteins constitute up to 30% of the human proteome. These proteins have special properties because the transmembrane segments are embedded into lipid bilayer while extramembranous parts are in different environments. Membrane proteins have several functions and are involved in numerous diseases. A large number of prediction methods have been introduced to predict protein subcellular localization as well as the tolerance or pathogenicity of amino acid substitutions.

Results

We tested the performance of 22 tolerance predictors by collecting information on membrane proteins and variants in them. The analysis indicated that the best tools had similar prediction performance on transmembrane, inside and outside regions of transmembrane proteins and comparable to overall prediction performances for all types of proteins. PON-P2 had the highest performance followed by REVEL, MetaSVM and VEST3. Further, we tested with the high quality dataset also the performance of seven subcellular localization predictors on membrane proteins. We assessed separately the performance for single pass and multi pass membrane proteins. Predictions for multi pass proteins were more reliable than those for single pass proteins.

Conclusions

The predictors for variant effects had better performance than subcellular localization tools. The best tolerance predictors are highly reliable. As there are large differences in the performances of tools, end-users have to be cautious in method selection.

Electronic supplementary material

The online version of this article (10.1186/s12864-019-5865-0) contains supplementary material, which is available to authorized users.

Topology prediction of insect olfactory receptors

Olfactory receptors are important transmembrane proteins that enable organisms to perceive odours and react to them. Structural understanding of insect olfactory receptors is scarce. In this review, we discuss different transmembrane helix prediction methods, consensus methods, topology prediction methods which can enable topology prediction of these proteins. We discuss the current success rates by applying the algorithms on few G-protein coupled receptors of known structure and olfactory receptor sequences and outstanding challenges. Finally, we discuss the impact of topology prediction on biology and modeling of ORs.

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.