A highly sensitive selection method for directed evolution of homing endonucleases

Intein-based thermoregulated meganucleases for containment of genetic material

Limiting the spread of synthetic genetic information outside of the intended use is essential for applications where biocontainment is critical. In particular, biocontainment of engineered probiotics and plasmids that are excreted from the mammalian gastrointestinal tract is needed to prevent escape and acquisition of genetic material that could confer a selective advantage to microbial communities. Here, we built a simple and lightweight biocontainment system that post-translationally activates a site-specific DNA endonuclease to degrade DNA at 18°C and not at higher temperatures. We constructed an orthogonal set of temperature-sensitive meganucleases (TSMs) by inserting the yeast VMA1 L212P temperature-sensitive intein into the coding regions of LAGLIDADG homing endonucleases. We showed that the TSMs eliminated plasmids carrying the cognate TSM target site from laboratory strains of Escherichia coli at the permissive 18°C but not at higher restrictive temperatures. Plasmid elimination is dependent on both TSM endonuclease activity and intein splicing. TSMs eliminated plasmids from E. coli Nissle 1917 after passage through the mouse gut when fecal resuspensions were incubated at 18°C but not at 37°C. Collectively, our data demonstrates the potential of thermoregulated meganucleases as a means of restricting engineered plasmids and probiotics to the mammalian gut.

Hydrolytic endonucleolytic ribozyme (HYER) is programmable for sequence-specific DNA cleavage

Ribozymes are catalytic RNAs with diverse functions including self-splicing and polymerization. This work aims to discover natural ribozymes that behave as hydrolytic and sequence-specific DNA endonucleases, which could be repurposed as DNA manipulation tools. Focused on bacterial group II-C introns, we found that many systems without intron-encoded protein propagate multiple copies in their resident genomes. These introns, named HYdrolytic Endonucleolytic Ribozymes (HYERs), cleaved RNA, single-stranded DNA, bubbled double-stranded DNA (dsDNA), and plasmids in vitro. HYER1 generated dsDNA breaks in the mammalian genome. Cryo-electron microscopy analysis revealed a homodimer structure for HYER1, where each monomer contains a Mg2+-dependent hydrolysis pocket and captures DNA complementary to the target recognition site (TRS). Rational designs including TRS extension, recruiting sequence insertion, and heterodimerization yielded engineered HYERs showing improved specificity and flexibility for DNA manipulation.

Friend or Foe: Protein Inhibitors of DNA Gyrase

DNA gyrase is essential for the successful replication of circular chromosomes, such as those found in most bacterial species, by relieving topological stressors associated with unwinding the double-stranded genetic material. This critical central role makes gyrase a valued target for antibacterial approaches, as exemplified by the highly successful fluoroquinolone class of antibiotics. It is reasonable that the activity of gyrase could be intrinsically regulated within cells, thereby helping to coordinate DNA replication with doubling times. Numerous proteins have been identified to exert inhibitory effects on DNA gyrase, although at lower doses, it can appear readily reversible and therefore may have regulatory value. Some of these, such as the small protein toxins found in plasmid-borne addiction modules, can promote cell death by inducing damage to DNA, resulting in an analogous outcome as quinolone antibiotics. Others, however, appear to transiently impact gyrase in a readily reversible and non-damaging mechanism, such as the plasmid-derived Qnr family of DNA-mimetic proteins. The current review examines the origins and known activities of protein inhibitors of gyrase and highlights opportunities to further exert control over bacterial growth by targeting this validated antibacterial target with novel molecular mechanisms. Furthermore, we are gaining new insights into fundamental regulatory strategies of gyrase that may prove important for understanding diverse growth strategies among different bacteria.

A Photoresponsive Homing Endonuclease for Programmed DNA Cleavage

Homing endonucleases are used in a wide range of biotechnological applications including gene editing, in gene drive systems, and for the modification of DNA structures, arrays, and prodrugs. However, controlling nuclease activity and sequence specificity remain key challenges when developing new tools. Here a photoresponsive homing endonuclease was engineered for optical control of DNA cleavage by partitioning DNA binding and nuclease domains of the monomeric homing endonuclease I-TevI into independent polypeptide chains. Use of the Aureochrome1a light-oxygen-voltage domain delivered control of dimerization with light. Illumination reduced the concentration needed to achieve 50% cleavage of the homing target site by 6-fold when compared to the dark state, resulting in an up to 9-fold difference in final yields between cleavage products. I-TevI nucleases with and without a native I-TevI zinc finger motif displayed different nuclease activity and sequence preference impacting the promiscuity of the nuclease domain. By harnessing an alternative DNA binding domain, target preference was reprogrammed only when the nuclease lacked the I-TevI zinc finger motif. This work establishes a first-generation photoresponsive platform for spatiotemporal activation of DNA cleavage.

Positive Selection Screens for Programmable Endonuclease Activity Using I-SceI

Positive selection screens are high-throughput assays to characterize novel enzymes from environmental samples and enrich for more powerful variants from libraries in applications such as biodiversity mining and directed evolution. However, overly stringent selection can limit the power of these screens due to a high false-negative rate. To create a more flexible and less restrictive screen for novel programmable DNA endonucleases, we developed a novel I-SceI-based platform. In this system, mutant E. coli genomes are cleaved upon induction of I-SceI to inhibit cell growth. Growth is rescued in an activity-dependent manner by plasmid curing or cleavage of the I-SceI expression plasmid via endonuclease candidates. More active candidates more readily proliferate and overtake growth of less active variants leading to enrichment. While demonstrated here with Cas9, this protocol can be readily adapted to any programmable DNA endonuclease and used to characterize single candidates or to enrich more powerful variants from pooled candidates or libraries.

Prokaryotic Argonautes for in vivo biotechnology and molecular diagnostics

Prokaryotic Argonautes (pAgos) are an emerging class of programmable endonucleases that are believed to be more flexible than existing CRISPR-Cas systems and have significant potential for biotechnology. Current applications of pAgos include a myriad of molecular diagnostics and in vitro DNA assembly tools. However, efforts have historically been centered on thermophilic pAgo variants. To enable in vivo biotechnological applications such as gene editing, focus has shifted to pAgos from mesophilic organisms. We discuss what is known of pAgos, how they are being developed for various applications, and strategies to overcome current challenges to in vivo applications in prokaryotes and eukaryotes.

A generalizable Cas9/sgRNA prediction model using machine transfer learning with small high-quality datasets

The CRISPR/Cas9 nuclease from Streptococcus pyogenes (SpCas9) can be used with single guide RNAs (sgRNAs) as a sequence-specific antimicrobial agent and as a genome-engineering tool. However, current bacterial sgRNA activity models struggle with accurate predictions and do not generalize well, possibly because the underlying datasets used to train the models do not accurately measure SpCas9/sgRNA activity and cannot distinguish on-target cleavage from toxicity. Here, we solve this problem by using a two-plasmid positive selection system to generate high-quality data that more accurately reports on SpCas9/sgRNA cleavage and that separates activity from toxicity. We develop a machine learning architecture (crisprHAL) that can be trained on existing datasets, that shows marked improvements in sgRNA activity prediction accuracy when transfer learning is used with small amounts of high-quality data, and that can generalize predictions to different bacteria. The crisprHAL model recapitulates known SpCas9/sgRNA-target DNA interactions and provides a pathway to a generalizable sgRNA bacterial activity prediction tool that will enable accurate antimicrobial and genome engineering applications.

Precise DNA cleavage using CRISPR-SpRYgests

Methods for in vitro DNA cleavage and molecular cloning remain unable to precisely cleave DNA directly adjacent to bases of interest. Restriction enzymes (REs) must bind specific motifs, whereas wild-type CRISPR-Cas9 or CRISPR-Cas12 nucleases require protospacer adjacent motifs (PAMs). Here we explore the utility of our previously reported near-PAMless SpCas9 variant, named SpRY, to serve as a universal DNA cleavage tool for various cloning applications. By performing SpRY DNA digests (SpRYgests) using more than 130 guide RNAs (gRNAs) sampling a wide diversity of PAMs, we discovered that SpRY is PAMless in vitro and can cleave DNA at practically any sequence, including sites refractory to cleavage with wild-type SpCas9. We illustrate the versatility and effectiveness of SpRYgests to improve the precision of several cloning workflows, including those not possible with REs or canonical CRISPR nucleases. We also optimize a rapid and simple one-pot gRNA synthesis protocol to streamline SpRYgest implementation. Together, SpRYgests can improve various DNA engineering applications that benefit from precise DNA breaks.

The compact Casπ (Cas12l) 'bracelet' provides a unique structural platform for DNA manipulation

CRISPR-Cas modules serve as the adaptive nucleic acid immune systems for prokaryotes, and provide versatile tools for nucleic acid manipulation in various organisms. Here, we discovered a new miniature type V system, CRISPR-Casπ (Cas12l) (~860 aa), from the environmental metagenome. Complexed with a large guide RNA (~170 nt) comprising the tracrRNA and crRNA, Casπ (Cas12l) recognizes a unique 5' C-rich PAM for DNA cleavage under a broad range of biochemical conditions, and generates gene editing in mammalian cells. Cryo-EM study reveals a 'bracelet' architecture of Casπ effector encircling the DNA target at 3.4 Å resolution, substantially different from the canonical 'two-lobe' architectures of Cas12 and Cas9 nucleases. The large guide RNA serves as a 'two-arm' scaffold for effector assembly. Our study expands the knowledge of DNA targeting mechanisms by CRISPR effectors, and offers an efficient but compact platform for DNA manipulation.

Characterization of a novel Escherichia coli recombineering selection/counterselection cassette

Recombineering is a highly efficient DNA cloning and modification technique by using the recombinase-mediated homologous recombination. Selection/counterselection cassette is often used in chromosomal DNA or large episomal DNA manipulation, in which the selection marker is used for the first step cassette selection while deleting the target gene via allelic exchange, and the counterselection marker is used for the second step replacement of the cassette by the foreign DNA fragment. A variety of selection/counterselection cassettes are reported, however, the cassettes suffer from the shortcomings of the requirement of pre-engineered strain or specific culture medium. Herein, we report a novel S-tetR- P_tetA-ccdB-aacC1-S selection/counterselection cassette that sidesteps the disadvantages. As a proof-of-concept, one-step gene cloning (0.7, 1.7, and 4.2 kb) and two-step Escherichia coli chromosomal gene knock-in (0.7 and 4.2 kb) were performed. The gene cloning and gene knock-in efficiencies are high up to 90%. The novel selection/counterselection cassette adds a powerful tool to the recombineering repertoire.

Improved genome editing by an engineered CRISPR-Cas12a

CRISPR-Cas12a is an RNA-guided, programmable genome editing enzyme found within bacterial adaptive immune pathways. Unlike CRISPR-Cas9, Cas12a uses only a single catalytic site to both cleave target double-stranded DNA (dsDNA) (cis-activity) and indiscriminately degrade single-stranded DNA (ssDNA) (trans-activity). To investigate how the relative potency of cis- versus trans-DNase activity affects Cas12a-mediated genome editing, we first used structure-guided engineering to generate variants of Lachnospiraceae bacterium Cas12a that selectively disrupt trans-activity. The resulting engineered mutant with the biggest differential between cis- and trans-DNase activity in vitro showed minimal genome editing activity in human cells, motivating a second set of experiments using directed evolution to generate additional mutants with robust genome editing activity. Notably, these engineered and evolved mutants had enhanced ability to induce homology-directed repair (HDR) editing by 2-18-fold compared to wild-type Cas12a when using HDR donors containing mismatches with crRNA at the PAM-distal region. Finally, a site-specific reversion mutation produced improved Cas12a (iCas12a) variants with superior genome editing efficiency at genomic sites that are difficult to edit using wild-type Cas12a. This strategy establishes a pipeline for creating improved genome editing tools by combining structural insights with randomization and selection. The available structures of other CRISPR-Cas enzymes will enable this strategy to be applied to improve the efficacy of other genome-editing proteins.

Repurposing the Homing Endonuclease I-SceI for Positive Selection and Development of Gene-Editing Technologies

Prokaryote genomes encode diverse programmable DNA endonucleases with significant potential for biotechnology and gene editing. However, these endonucleases differ significantly in their properties, which must be screened and measured. While positive selection screens based on ccdB and barnase have been developed to evaluate such proteins, their high levels of toxicity make them challenging to use. Here, we develop and validate a more robust positive selection screen based on the homing endonuclease I-SceI. Candidate endonucleases target and cure the I-SceI expression plasmid preventing induction of I-SceI-mediated double strand DNA breaks that lead to cell death in E. coli. We validated this screen to measure the relative activity of SpCas9, xCas9, and eSpCas9 and demonstrated an ability to enrich for more active endonuclease variants from a mixed population. This system may be applied in high throughput to rapidly characterize novel programmable endonucleases and be adapted for directed evolution of endonuclease function.

Harnessing the power of directed evolution to improve genome editing systems

The recent development of genome editing systems, such as zinc-finger nucleases, transcription activator-like effectors, CRISPR-Cas nucleases, and base editors has enabled the unprecedented capability to engineer the genomes of living cells. The ever-increasing demand for genome editors with improved accuracy, activity, and functionality has stimulated significant efforts to further engineer the genome editing systems. Directed evolution represents a promising strategy to improve the existing genome editing systems and enable new editing functions. Here, we review recent representative strategies to harness the power of directed evolution to improve genome editing systems, which have led to state-of-the-art genome editors that have significant implications for diverse applications in both laboratories and clinics.

Exploring the fitness benefits of genome reduction in Escherichia coli by a selection-driven approach

Artificial simplification of bacterial genomes is thought to have the potential to yield cells with reduced complexity, enhanced genetic stability, and improved cellular economy. Of these goals, economical gains, supposedly due to the elimination of superfluous genetic material, and manifested in elevated growth parameters in selected niches, have not yet been convincingly achieved. This failure might stem from limitations of the targeted genome reduction approach that assumes full knowledge of gene functions and interactions, and allows only a limited number of reduction trajectories to interrogate. To explore the potential fitness benefits of genome reduction, we generated successive random deletions in E. coli by a novel, selection-driven, iterative streamlining process. The approach allows the exploration of multiple streamlining trajectories, and growth periods inherent in the procedure ensure selection of the fittest variants of the population. By generating single- and multiple-deletion strains and reconstructing the deletions in the parental genetic background, we showed that favourable deletions can be obtained and accumulated by the procedure. The most reduced multiple-deletion strain, obtained in five deletion cycles (2.5% genome reduction), outcompeted the wild-type, and showed elevated biomass yield. The spectrum of advantageous deletions, however, affecting only a few genomic regions, appears to be limited.

A genome-wide transcriptomic analysis of protein-coding genes in human blood cells

Blood is the predominant source for molecular analyses in humans, both in clinical and research settings. It is the target for many therapeutic strategies, emphasizing the need for comprehensive molecular maps of the cells constituting human blood. In this study, we performed a genome-wide transcriptomic analysis of protein-coding genes in sorted blood immune cell populations to characterize the expression levels of each individual gene across the blood cell types. All data are presented in an interactive, open-access Blood Atlas as part of the Human Protein Atlas and are integrated with expression profiles across all major tissues to provide spatial classification of all protein-coding genes. This allows for a genome-wide exploration of the expression profiles across human immune cell populations and all major human tissues and organs.

Modifying a covarying protein-DNA interaction changes substrate preference of a site-specific endonuclease

Identifying and validating intermolecular covariation between proteins and their DNA-binding sites can provide insights into mechanisms that regulate selectivity and starting points for engineering new specificity. LAGLIDADG homing endonucleases (meganucleases) can be engineered to bind non-native target sites for gene-editing applications, but not all redesigns successfully reprogram specificity. To gain a global overview of residues that influence meganuclease specificity, we used information theory to identify protein-DNA covariation. Directed evolution experiments of one predicted pair, 227/+3, revealed variants with surprising shifts in I-OnuI substrate preference at the central 4 bases where cleavage occurs. Structural studies showed significant remodeling distant from the covarying position, including restructuring of an inter-hairpin loop, DNA distortions near the scissile phosphates, and new base-specific contacts. Our findings are consistent with a model whereby the functional impacts of covariation can be indirectly propagated to neighboring residues outside of direct contact range, allowing meganucleases to adapt to target site variation and indirectly expand the sequence space accessible for cleavage. We suggest that some engineered meganucleases may have unexpected cleavage profiles that were not rationally incorporated during the design process.

Active site residue identity regulates cleavage preference of LAGLIDADG homing endonucleases

LAGLIDADG homing endonucleases (meganucleases) are site-specific mobile endonucleases that can be adapted for genome-editing applications. However, one problem when reprogramming meganucleases on non-native substrates is indirect readout of DNA shape and flexibility at the central 4 bases where cleavage occurs. To understand how the meganuclease active site regulates DNA cleavage, we used functional selections and deep sequencing to profile the fitness landscape of 1600 I-LtrI and I-OnuI active site variants individually challenged with 67 substrates with central 4 base substitutions. The wild-type active site was not optimal for cleavage on many substrates, including the native I-LtrI and I-OnuI targets. Novel combinations of active site residues not observed in known meganucleases supported activity on substrates poorly cleaved by the wild-type enzymes. Strikingly, combinations of E or D substitutions in the two metal-binding residues greatly influenced cleavage activity, and E184D variants had a broadened cleavage profile. Analyses of I-LtrI E184D and the wild-type proteins co-crystallized with the non-cognate AACC central 4 sequence revealed structural differences that correlated with kinetic constants for cleavage of individual DNA strands. Optimizing meganuclease active sites to enhance cleavage of non-native central 4 target sites is a straightforward addition to engineering workflows that will expand genome-editing applications.

Diversifying the structure of zinc finger nucleases for high-precision genome editing

Genome editing for therapeutic applications often requires cleavage within a narrow sequence window. Here, to enable such high-precision targeting with zinc-finger nucleases (ZFNs), we have developed an expanded set of architectures that collectively increase the configurational options available for design by a factor of 64. These new architectures feature the functional attachment of the FokI cleavage domain to the amino terminus of one or both zinc-finger proteins (ZFPs) in the ZFN dimer, as well as the option to skip bases between the target triplets of otherwise adjacent fingers in each zinc-finger array. Using our new architectures, we demonstrate targeting of an arbitrarily chosen 28 bp genomic locus at a density that approaches 1.0 (i.e., efficient ZFNs available for targeting almost every base step). We show that these new architectures may be used for targeting three loci of therapeutic significance with a high degree of precision, efficiency, and specificity.

Directed evolution studies of a thermophilic Type II-C Cas9

Though making up nearly half of the known CRISPR-Cas9 family of enzymes, the Type II-C CRISPR-Cas9 has been underexplored for their molecular mechanisms and potential in safe gene editing applications. In comparison with the more popular Type II-A CRISPR-Cas9, the Type II-C enzymes are generally smaller in size and utilize longer base pairing in identification of their DNA substrates. These characteristics suggest easier portability and potentially less off-targets for Type II-C in gene editing applications. We describe identification and biochemical characterization of a thermophilic Type II-C CRISPR-Cas from Acidothermus cellulolyticus (AceCas9). We describe several library-based methods that enabled us to identify the PAM sequence and elements critical to protospacer mismatch surveillance of AceCas9.

Phosphate Lock Residues of Acidothermus cellulolyticus Cas9 Are Critical to Its Substrate Specificity

Despite being utilized widely in genome sciences, CRISPR-Cas9 remains limited in achieving high fidelity in cleaving DNA. A better understanding of the molecular basis of Cas9 holds the key to improve Cas9-based tools. We employed direct evolution and in vitro characterizations to explore structural parameters that impact the specificity of the thermophilic Cas9 from Acidothermus cellulolyticus (AceCas9). By identifying variants that are able to cleave mismatched protospacers within the seed region, we found a critical role of the phosphate lock residues in substrate specificity in a manner that depends on their sizes and charges. Removal of the negative charge from the phosphate lock residues significantly decreases sensitivity to the guide-DNA mismatches. An increase in size of the substituted residues further reduces the sensitivity to mismatches at the first position of the protospacer. Our findings identify the phosphate lock residues as an important site for tuning the specificity and catalytic efficiency of Cas9.

Genetic Analysis of Rare Human Variants of Regulators of G Protein Signaling Proteins and Their Role in Human Physiology and Disease

Regulators of G protein signaling (RGS) proteins modulate the physiologic actions of many neurotransmitters, hormones, and other signaling molecules. Human RGS proteins comprise a family of 20 canonical proteins that bind directly to G protein-coupled receptors/G protein complexes to limit the lifetime of their signaling events, which regulate all aspects of cell and organ physiology. Genetic variations account for diverse human traits and individual predispositions to disease. RGS proteins contribute to many complex polygenic human traits and pathologies such as hypertension, atherosclerosis, schizophrenia, depression, addiction, cancers, and many others. Recent analysis indicates that most human diseases are due to extremely rare genetic variants. In this study, we summarize physiologic roles for RGS proteins and links to human diseases/traits and report rare variants found within each human RGS protein exome sequence derived from global population studies. Each RGS sequence is analyzed using recently described bioinformatics and proteomic tools for measures of missense tolerance ratio paired with combined annotation-dependent depletion scores, and protein post-translational modification (PTM) alignment cluster analysis. We highlight selected variants within the well-studied RGS domain that likely disrupt RGS protein functions and provide comprehensive variant and PTM data for each RGS protein for future study. We propose that rare variants in functionally sensitive regions of RGS proteins confer profound change-of-function phenotypes that may contribute, in newly appreciated ways, to complex human diseases and/or traits. This information provides investigators with a valuable database to explore variation in RGS protein function, and for targeting RGS proteins as future therapeutic targets.

Repurposing CRISPR-Cas12b for mammalian genome engineering

The prokaryotic CRISPR-Cas adaptive immune systems provide valuable resources to develop genome editing tools, such as CRISPR-Cas9 and CRISPR-Cas12a/Cpf1. Recently, CRISPR-Cas12b/C2c1, a distinct type V-B system, has been characterized as a dual-RNA-guided DNA endonuclease system. Though being active in vitro, its cleavage activity at endogenous genome remains to be explored. Furthermore, the optimal cleavage temperature of the reported Cas12b orthologs is higher than 40 °C, which is unsuitable for mammalian applications. Here, we report the identification of a Cas12b system from the Alicyclobacillus acidiphilus (AaCas12b), which maintains optimal nuclease activity over a wide temperature range (31 °C–59 °C). AaCas12b can be repurposed to engineer mammalian genomes for versatile applications, including single and multiplex genome editing, gene activation, and generation of gene mutant mouse models. Moreover, whole-genome sequencing reveals high specificity and minimal off-target effects of AaCas12b-meditated genome editing. Our findings establish CRISPR-Cas12b as a versatile tool for mammalian genome engineering.

Directed evolution of CRISPR-Cas9 to increase its specificity

The use of CRISPR-Cas9 as a therapeutic reagent is hampered by its off-target effects. Although rationally designed S. pyogenes Cas9 (SpCas9) variants that display higher specificities than the wild-type SpCas9 protein are available, these attenuated Cas9 variants are often poorly efficient in human cells. Here, we develop a directed evolution approach in E. coli to obtain Sniper-Cas9, which shows high specificities without killing on-target activities in human cells. Unlike other engineered Cas9 variants, Sniper-Cas9 shows WT-level on-target activities with extended or truncated sgRNAs with further reduced off-target activities and works well in a preassembled ribonucleoprotein (RNP) format to allow DNA-free genome editing.

A high-fidelity Cas9 mutant delivered as a ribonucleoprotein complex enables efficient gene editing in human hematopoietic stem and progenitor cells

Translation of the CRISPR-Cas9 system to human therapeutics holds high promise. However, specificity remains a concern especially when modifying stem cell populations. We show that existing rationally engineered Cas9 high-fidelity variants have reduced on-target activity when using the therapeutically relevant ribonucleoprotein (RNP) delivery method. Therefore, we devised an unbiased bacterial screen to isolate variants that retain activity in the RNP format. Introduction of a single point mutation, p.R691A, in Cas9 (high-fidelity (HiFi) Cas9) retained the high on-target activity of Cas9 while reducing off-target editing. HiFi Cas9 induces robust AAV6-mediated gene targeting at five therapeutically relevant loci (HBB, IL2RG, CCR5, HEXB, and TRAC) in human CD34+ hematopoietic stem and progenitor cells (HSPCs) as well as primary T cells. We also show that HiFi Cas9 mediates high-level correction of the sickle cell disease (SCD)-causing p.E6V mutation in HSPCs derived from patients with SCD. We anticipate that HiFi Cas9 will have wide utility for both basic science and therapeutic genome-editing applications.

The Impact of DNA Topology and Guide Length on Target Selection by a Cytosine-Specific Cas9

Cas9 is an RNA-guided DNA cleavage enzyme being actively developed for genome editing and gene regulation. To be cleaved by Cas9, a double stranded DNA, or the protospacer, must be complementary to the guide region, typically 20-nucleotides in length, of the Cas9-bound guide RNA, and adjacent to a short Cas9-specific element called Protospacer Adjacent Motif (PAM). Understanding the correct juxtaposition of the protospacer- and PAM-interaction with Cas9 will enable development of versatile and safe Cas9-based technology. We report identification and biochemical characterization of Cas9 from Acidothermus cellulolyticus (AceCas9). AceCas9 depends on a 5'-NNNCC-3' PAM and is more efficient in cleaving negative supercoils than relaxed DNA. Kinetic as well as in vivo activity assays reveal that AceCas9 achieves optimal activity when combined with a guide RNA containing a 24-nucleotide complementarity region. The cytosine-specific, DNA topology-sensitive, and extended guide-dependent properties of AceCas9 may be explored for specific genome editing applications.

Towards improved genome-scale metabolic network reconstructions: unification, transcript specificity and beyond

Genome-scale metabolic network reconstructions provide a basis for the investigation of the metabolic properties of an organism. There are reconstructions available for multiple organisms, from prokaryotes to higher organisms and methods for the analysis of a reconstruction. One example is the use of flux balance analysis to improve the yields of a target chemical, which has been applied successfully. However, comparison of results between existing reconstructions and models presents a challenge because of the heterogeneity of the available reconstructions, for example, of standards for presenting gene-protein-reaction associations, nomenclature of metabolites and reactions or selection of protonation states. The lack of comparability for gene identifiers or model-specific reactions without annotated evidence often leads to the creation of a new model from scratch, as data cannot be properly matched otherwise. In this contribution, we propose to improve the predictive power of metabolic models by switching from gene-protein-reaction associations to transcript-isoform-reaction associations, thus taking advantage of the improvement of precision in gene expression measurements. To achieve this precision, we discuss available databases that can be used to retrieve this type of information and point at issues that can arise from their neglect. Further, we stress issues that arise from non-standardized building pipelines, like inconsistencies in protonation states. In addition, problems arising from the use of non-specific cofactors, e.g. artificial futile cycles, are discussed, and finally efforts of the metabolic modelling community to unify model reconstructions are highlighted.

Protein Data Bank Japan (PDBj): updated user interfaces, resource description framework, analysis tools for large structures

The Protein Data Bank Japan (PDBj, http://pdbj.org), a member of the worldwide Protein Data Bank (wwPDB), accepts and processes the deposited data of experimentally determined macromolecular structures. While maintaining the archive in collaboration with other wwPDB partners, PDBj also provides a wide range of services and tools for analyzing structures and functions of proteins. We herein outline the updated web user interfaces together with RESTful web services and the backend relational database that support the former. To enhance the interoperability of the PDB data, we have previously developed PDB/RDF, PDB data in the Resource Description Framework (RDF) format, which is now a wwPDB standard called wwPDB/RDF. We have enhanced the connectivity of the wwPDB/RDF data by incorporating various external data resources. Services for searching, comparing and analyzing the ever-increasing large structures determined by hybrid methods are also described.

DnaK-Dependent Accelerated Evolutionary Rate in Prokaryotes

Many proteins depend on an interaction with molecular chaperones in order to fold into a functional tertiary structure. Previous studies showed that protein interaction with the GroEL/GroES chaperonine and Hsp90 chaperone can buffer the impact of slightly deleterious mutations in the protein sequence. This capacity of GroEL/GroES to prevent protein misfolding has been shown to accelerate the evolution of its client proteins. Whether other bacterial chaperones have a similar effect on their client proteins is currently unknown. Here, we study the impact of DnaK (Hsp70) chaperone on the evolution of its client proteins. Evolutionary parameters were derived from comparison of the Escherichia coli proteome to 1,808,565 orthologous proteins in 1,149 proteobacterial genomes. Our analysis reveals a significant positive correlation between protein binding frequency with DnaK and evolutionary rate. Proteins with high binding affinity to DnaK evolve on average 4.3-fold faster than proteins in the lowest binding affinity class at the genus resolution. Differences in evolutionary rates of DnaK interactor classes are still significant after adjusting for possible effects caused by protein expression level. Furthermore, we observe an additive effect of DnaK and GroEL chaperones on the evolutionary rates of their common interactors. Finally, we found pronounced similarities in the physicochemical profiles that characterize proteins belonging to DnaK and GroEL interactomes. Our results thus implicate DnaK-mediated folding as a major component in shaping protein evolutionary dynamics in bacteria and supply further evidence for the long-term manifestation of chaperone-mediated folding on genome evolution.

MMseqs software suite for fast and deep clustering and searching of large protein sequence sets

MOTIVATION

Sequence databases are growing fast, challenging existing analysis pipelines. Reducing the redundancy of sequence databases by similarity clustering improves speed and sensitivity of iterative searches. But existing tools cannot efficiently cluster databases of the size of UniProt to 50% maximum pairwise sequence identity or below. Furthermore, in metagenomics experiments typically large fractions of reads cannot be matched to any known sequence anymore because searching with sensitive but relatively slow tools (e.g. BLAST or HMMER3) through comprehensive databases such as UniProt is becoming too costly.

RESULTS

MMseqs (Many-against-Many sequence searching) is a software suite for fast and deep clustering and searching of large datasets, such as UniProt, or 6-frame translated metagenomics sequencing reads. MMseqs contains three core modules: a fast and sensitive prefiltering module that sums up the scores of similar k-mers between query and target sequences, an SSE2- and multi-core-parallelized local alignment module, and a clustering module.In our homology detection benchmarks, MMseqs is much more sensitive and 4-30 times faster than UBLAST and RAPsearch, respectively, although it does not reach BLAST sensitivity yet. Using its cascaded clustering workflow, MMseqs can cluster large databases down to ∼30% sequence identity at hundreds of times the speed of BLASTclust and much deeper than CD-HIT and USEARCH. MMseqs can also update a database clustering in linear instead of quadratic time. Its much improved sensitivity-speed trade-off should make MMseqs attractive for a wide range of large-scale sequence analysis tasks.

AVAILABILITY AND IMPLEMENTATION

MMseqs is open-source software available under GPL at https://github.com/soedinglab/MMseqs

CONTACT

martin.steinegger@mpibpc.mpg.de, soeding@mpibpc.mpg.de

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Broadening the targeting range of Staphylococcus aureus CRISPR-Cas9 by modifying PAM recognition

CRISPR-Cas9 nucleases target specific DNA sequences using a guide RNA but also require recognition of a protospacer adjacent motif (PAM) by the Cas9 protein. Although longer PAMs can potentially improve the specificity of genome editing, they limit the range of sequences that Cas9 orthologs can target. One potential strategy to relieve this restriction is to relax the PAM recognition specificity of Cas9. Here we used molecular evolution to modify the NNGRRT PAM of Staphylococcus aureus Cas9 (SaCas9). One variant we identified, referred to as KKH SaCas9, showed robust genome editing activities at endogenous human target sites with NNNRRT PAMs, thereby increasing SaCas9 targeting range by two- to fourfold. Using GUIDE-seq, we show that wild-type and KKH SaCas9 induce comparable numbers of off-target effects in human cells. Our strategy for evolving PAM specificity does not require structural information and therefore should be applicable to a wide range of Cas9 orthologs.

Connecting proteins with drug-like compounds: Open source drug discovery workflows with BindingDB and KNIME

Today's large, public databases of protein-small molecule interaction data are creating important new opportunities for data mining and integration. At the same time, new graphical user interface-based workflow tools offer facile alternatives to custom scripting for informatics and data analysis. Here, we illustrate how the large protein-ligand database BindingDB may be incorporated into KNIME workflows as a step toward the integration of pharmacological data with broader biomolecular analyses. Thus, we describe a collection of KNIME workflows that access BindingDB data via RESTful webservices and, for more intensive queries, via a local distillation of the full BindingDB dataset. We focus in particular on the KNIME implementation of knowledge-based tools to generate informed hypotheses regarding protein targets of bioactive compounds, based on notions of chemical similarity. A number of variants of this basic approach are tested for seven existing drugs with relatively ill-defined therapeutic targets, leading to replication of some previously confirmed results and discovery of new, high-quality hits. Implications for future development are discussed. Database URL: www.bindingdb.org.

SInCRe-structural interactome computational resource for Mycobacterium tuberculosis

We have developed an integrated database for Mycobacterium tuberculosis H37Rv (Mtb) that collates information on protein sequences, domain assignments, functional annotation and 3D structural information along with protein–protein and protein–small molecule interactions. SInCRe (Structural Interactome Computational Resource) is developed out of CamBan (Cambridge and Bangalore) collaboration. The motivation for development of this database is to provide an integrated platform to allow easily access and interpretation of data and results obtained by all the groups in CamBan in the field of Mtb informatics. In-house algorithms and databases developed independently by various academic groups in CamBan are used to generate Mtb-specific datasets and are integrated in this database to provide a structural dimension to studies on tuberculosis. The SInCRe database readily provides information on identification of functional domains, genome-scale modelling of structures of Mtb proteins and characterization of the small-molecule binding sites within Mtb. The resource also provides structure-based function annotation, information on small-molecule binders including FDA (Food and Drug Administration)-approved drugs, protein–protein interactions (PPIs) and natural compounds that bind to pathogen proteins potentially and result in weakening or elimination of host–pathogen protein–protein interactions. Together they provide prerequisites for identification of off-target binding.

Database URL:http://proline.biochem.iisc.ernet.in/sincre

Engineered CRISPR-Cas9 nucleases with altered PAM specificities

Although CRISPR-Cas9 nucleases are widely used for genome editing^{1, 2}, the range of sequences that Cas9 can recognize is constrained by the need for a specific protospacer adjacent motif (PAM)^3–6. As a result, it can often be difficult to target double-stranded breaks (DSBs) with the precision that is necessary for various genome editing applications. The ability to engineer Cas9 derivatives with purposefully altered PAM specificities would address this limitation. Here we show that the commonly used Streptococcus pyogenes Cas9 (SpCas9) can be modified to recognize alternative PAM sequences using structural information, bacterial selection-based directed evolution, and combinatorial design. These altered PAM specificity variants enable robust editing of endogenous gene sites in zebrafish and human cells not currently targetable by wild-type SpCas9, and their genome-wide specificities are comparable to wild-type SpCas9 as judged by GUIDE-Seq analysis⁷. In addition, we identified and characterized another SpCas9 variant that exhibits improved specificity in human cells, possessing better discrimination against off-target sites with non-canonical NAG and NGA PAMs and/or mismatched spacers. We also found that two smaller-size Cas9 orthologues, Streptococcus thermophilus Cas9 (St1Cas9) and Staphylococcus aureus Cas9 (SaCas9), function efficiently in the bacterial selection systems and in human cells, suggesting that our engineering strategies could be extended to Cas9s from other species. Our findings provide broadly useful SpCas9 variants and, more importantly, establish the feasibility of engineering a wide range of Cas9s with altered and improved PAM specificities.

Pathways with PathWhiz

PathWhiz (http://smpdb.ca/pathwhiz) is a web server designed to create colourful, visually pleasing and biologically accurate pathway diagrams that are both machine-readable and interactive. As a web server, PathWhiz is accessible from almost any place and compatible with essentially any operating system. It also houses a public library of pathways and pathway components that can be easily viewed and expanded upon by its users. PathWhiz allows users to readily generate biologically complex pathways by using a specially designed drawing palette to quickly render metabolites (including automated structure generation), proteins (including quaternary structures, covalent modifications and cofactors), nucleic acids, membranes, subcellular structures, cells, tissues and organs. Both small-molecule and protein/gene pathways can be constructed by combining multiple pathway processes such as reactions, interactions, binding events and transport activities. PathWhiz's pathway replication and propagation functions allow for existing pathways to be used to create new pathways or for existing pathways to be automatically propagated across species. PathWhiz pathways can be saved in BioPAX, SBGN-ML and SBML data exchange formats, as well as PNG, PWML, HTML image map or SVG images that can be viewed offline or explored using PathWhiz's interactive viewer. PathWhiz has been used to generate over 700 pathway diagrams for a number of popular databases including HMDB, DrugBank and SMPDB.

Control of catalytic efficiency by a coevolving network of catalytic and noncatalytic residues

The active sites of enzymes consist of residues necessary for catalysis and structurally important noncatalytic residues that together maintain the architecture and function of the active site. Examples of evolutionary interactions between catalytic and noncatalytic residues have been difficult to define and experimentally validate due to a general intolerance of these residues to substitution. Here, using computational methods to predict coevolving residues, we identify a network of positions consisting of two catalytic metal-binding residues and two adjacent noncatalytic residues in LAGLIDADG homing endonucleases (LHEs). Distinct combinations of the four residues in the network map to distinct LHE subfamilies, with a striking distribution of the metal-binding Asp (D) and Glu (E) residues. Mutation of these four positions in three LHEs--I-LtrI, I-OnuI, and I-HjeMI--indicate that the combinations of residues tolerated are specific to each enzyme. Kinetic analyses under single-turnover conditions revealed that I-LtrI activity could be modulated over an ∼100-fold range by mutation of residues in the coevolving network. I-LtrI catalytic site variants with low activity could be rescued by compensatory mutations at adjacent noncatalytic sites that restore an optimal coevolving network and vice versa. Our results demonstrate that LHE activity is constrained by an evolutionary barrier of residues with strong context-dependent effects. Creation of optimal coevolving active-site networks is therefore an important consideration in engineering of LHEs and other enzymes.

I-OmiI and I-OmiII: two intron-encoded homing endonucleases within the Ophiostoma minus rns gene

The mitochondrial small subunit ribosomal RNA (rns) gene of the ascomycetous fungus Ophiostoma minus [strain WIN(M)371] was found to contain a group IC2 and a group IIB1 intron at positions mS569 and mS952 respectively. Both introns have open reading frames (ORFs) embedded that encode double motif LAGLIDADG homing endonucleases (I-OmiI and I-OmiII respectively). Codon-optimized versions of I-OmiI and I-OmiII were synthesized for overexpression in Escherichia coli. The in vitro characterization of I-OmiII showed that it is a functional homing endonuclease that cleaves the rns target site two nucleotides upstream (sense strand) of the intron insertion site generating 4 nucleotide 3' overhangs. The endonuclease activity of I-OmiII was tested using linear and circular substrates and cleavage activity was evaluated at various temperatures. The I-OmiI protein was expressed in E. coli, but purification was difficult, thus the endonuclease activity of this protein was tested via in vivo assays. Overall this study showed that there are many native forms of functional homing endonucleases yet to be discovered among fungal mtDNA genomes.

TuberQ: a Mycobacterium tuberculosis protein druggability database

In 2012 an estimated 8.6 million people developed tuberculosis (TB) and 1.3 million died from the disease [including 320 000 deaths among human immunodeficiency virus (HIV)-positive people]. There is an urgent need for new anti-TB drugs owing to the following: the fact that current treatments have severe side effects, the increasing emergence of multidrug-resistant strains of Mycobacterium tuberculosis (Mtb), the negative drug-drug interactions with certain HIV (or other disease) treatments and the ineffectiveness against dormant Mtb. In this context we present here the TuberQ database, a novel resource for all researchers working in the field of drug development in TB. The main feature of TuberQ is to provide a druggability analysis of Mtb proteins in a consistent and effective manner, contributing to a better selection of potential drug targets for screening campaigns and the analysis of targets for structure-based drug design projects. The structural druggability analysis is combined with features related to the characteristics of putative inhibitor binding pockets and with functional and biological data of proteins. The structural analysis is performed on all available unique Mtb structures and high-quality structural homology-based models. This information is shown in an interactive manner, depicting the protein structure, the pockets and the associated characteristics for each protein. TuberQ also provides information about gene essentiality information, as determined from whole cell-based knockout experiments, and expression information obtained from microarray experiments done in different stress-related conditions. We hope that TuberQ will be a powerful tool for researchers working in TB and eventually will lead to the identification of novel putative targets and progresses in therapeutic activities. Database URL: http://tuberq.proteinq.com.ar/

A two-plasmid bacterial selection system for characterization and engineering of homing endonucleases

Homing endonucleases recognize long DNA sequences and generate site-specific DNA double-stranded breaks. They can serve as a powerful genomic modification tool in various industrial and biomedical applications. Here, we describe a two-plasmid bacterial selection system for characterization and engineering of homing endonucleases. This selection system couples the DNA cleavage activity of a homing endonuclease with the survival of host cells. Therefore, it can be used for assaying in vivo activity of homing endonucleases. Moreover, due to its high sensitivity, it can be applied for directed evolution of homing endonucleases with altered sequence specificity.

Rapid screening of endonuclease target site preference using a modified bacterial two-plasmid selection

Homing endonucleases and other site-specific endonucleases have potential applications in genome editing, yet efficient targeting requires a thorough understanding of DNA-sequence specificity. Here, we describe a modified two-plasmid genetic selection in Escherichia coli that allows rapid profiling of nucleotide substitutions within a target site of given endonucleases. The selection utilizes a toxic plasmid (pTox) that encodes a DNA gyrase toxin in addition to the endonuclease target site. Cleavage of the toxic plasmid by an endonuclease expressed from a second plasmid (pEndo) facilitates growth under selective conditions. The modified protocol utilizes competent cells harboring the endonuclease expression plasmid into which target site plasmids are transformed. Replica plating on nonselective and selective media plates identifies cleavable and non-cleavable targets. Thus, a library of randomized target sites, or many individual target sites, can be analyzed using a single transformation. Both cleavable and non-cleavable targets can be analyzed by DNA sequencing to gain information about nucleotide preference in the endonuclease's target site.

A yeast-based recombination assay for homing endonuclease activity

Homing endonucleases (HEs) are natural enzymes that cleave long DNA target with a high specificity and trigger homologous recombination at the exact site of the break. Such mechanisms can thus be used for all the applications covered today by the generic name of "genome engineering": targeted sequence insertion, removal, or editing. However, before being able to address those applications, the engineering of HEs must be mastered so that any potential target would be efficiently and specifically recognized and cleaved. Working on the I-CreI model, we have developed a very powerful platform to generate HEs with new tailored specificity. We have put in place the first in vivo, functional, high throughput assay to generate I-CreI variants and measure their activity. We use semi-rational design combined with proprietary in silico predictions to design and synthesize I-CreI mutants that are tested for their capacity to induce homologous recombination in a yeast cell. The process has been standardized and robotized so that we can generate thousands of I-CreI derivatives, characterize their cleavage profile, and deliver them for further applications in the research, therapeutic, or agrobusiness fields.

Determining the specificities of TALENs, Cas9, and other genome-editing enzymes

The rapid development of programmable site-specific endonucleases has led to a dramatic increase in genome engineering activities for research and therapeutic purposes. Specific loci of interest in the genomes of a wide range of organisms including mammals can now be modified using zinc-finger nucleases, transcription activator-like effectornucleases, and CRISPR-associated Cas9 endonucleases in a site-specific manner, in some cases requiring relatively modest effort for endonuclease design, construction, and application. While these technologies have made genome engineering widely accessible, the ability of programmable nucleases to cleave off-target sequences can limit their applicability and raise concerns about therapeutic safety. In this chapter, we review methods to evaluate and improve the DNA cleavage activity of programmable site-specific endonucleases and describe a procedure for a comprehensive off-target profiling method based on the in vitro selection of very large (~10(12)-membered) libraries of potential nuclease substrates.

A bacterial one-hybrid system to isolate homing endonuclease variants with altered DNA target specificities

Chromosomal cleavage near the site of mutations that cause disease can facilitate the targeted repair of the locus. Gene therapy protocols therefore require the engineering of DNA endonucleases that target specific genomic loci. Here, we describe a bacterial one-hybrid selection system that has been used to isolate derivatives of the I-SceI homing endonuclease from combinatorial libraries that display altered DNA recognition specificities. The construction of plasmid expression libraries, the development of reporter strains, and the utilization of these components in the bacterial one-hybrid system are detailed.

ModBase, a database of annotated comparative protein structure models and associated resources

ModBase (http://salilab.org/modbase) is a database of annotated comparative protein structure models. The models are calculated by ModPipe, an automated modeling pipeline that relies primarily on Modeller for fold assignment, sequence-structure alignment, model building and model assessment (http://salilab.org/modeller/). ModBase currently contains almost 30 million reliable models for domains in 4.7 million unique protein sequences. ModBase allows users to compute or update comparative models on demand, through an interface to the ModWeb modeling server (http://salilab.org/modweb). ModBase models are also available through the Protein Model Portal (http://www.proteinmodelportal.org/). Recently developed associated resources include the AllosMod server for modeling ligand-induced protein dynamics (http://salilab.org/allosmod), the AllosMod-FoXS server for predicting a structural ensemble that fits an SAXS profile (http://salilab.org/allosmod-foxs), the FoXSDock server for protein–protein docking filtered by an SAXS profile (http://salilab.org/foxsdock), the SAXS Merge server for automatic merging of SAXS profiles (http://salilab.org/saxsmerge) and the Pose & Rank server for scoring protein–ligand complexes (http://salilab.org/poseandrank). In this update, we also highlight two applications of ModBase: a PSI:Biology initiative to maximize the structural coverage of the human alpha-helical transmembrane proteome and a determination of structural determinants of human immunodeficiency virus-1 protease specificity.

Germline excision of transgenes in Aedes aegypti by homing endonucleases

Aedes (Ae.) aegypti is the primary vector for dengue viruses (serotypes1–4) and chikungunya virus. Homing endonucleases (HEs) are ancient selfish elements that catalyze double-stranded DNA breaks (DSB) in a highly specific manner. In this report, we show that the HEs Y2-I-AniI, I-CreI and I-SceI are all capable of catalyzing the excision of genomic segments from the Ae. aegypti genome in a heritable manner. Y2-I-AniI demonstrated the highest efficiency at two independent genomic targets, with 20–40% of Y2-I-AniI-treated individuals producing offspring that had lost the target transgene. HE-induced DSBs were found to be repaired via the single-strand annealing (SSA) and non-homologous end-joining (NHEJ) pathways in a manner dependent on the availability of direct repeat sequences in the transgene. These results support the development of HE-based gene editing and gene drive strategies in Ae. aegypti, and confirm the utility of HEs in the manipulation and modification of transgenes in this important vector.

Explaining microbial phenotypes on a genomic scale: GWAS for microbes

There is an increasing availability of complete or draft genome sequences for microbial organisms. These data form a potentially valuable resource for genotype-phenotype association and gene function prediction, provided that phenotypes are consistently annotated for all the sequenced strains. In this review, we address the requirements for successful gene-trait matching. We outline a basic protocol for microbial functional genomics, including genome assembly, annotation of genotypes (including single nucleotide polymorphisms, orthologous groups and prophages), data pre-processing, genotype-phenotype association, visualization and interpretation of results. The methodologies for association described herein can be applied to other data types, opening up possibilities to analyze transcriptome-phenotype associations, and correlate microbial population structure or activity, as measured by metagenomics, to environmental parameters.

The monomeric GIY-YIG homing endonuclease I-BmoI uses a molecular anchor and a flexible tether to sequentially nick DNA

The GIY-YIG nuclease domain is found within protein scaffolds that participate in diverse cellular pathways and contains a single active site that hydrolyzes DNA by a one-metal ion mechanism. GIY-YIG homing endonucleases (GIY-HEs) are two-domain proteins with N-terminal GIY-YIG nuclease domains connected to C-terminal DNA-binding and they are thought to function as monomers. Using I-BmoI as a model GIY-HE, we test mechanisms by which the single active site is used to generate a double-strand break. We show that I-BmoI is partially disordered in the absence of substrate, and that the GIY-YIG domain alone has weak affinity for DNA. Significantly, we show that I-BmoI functions as a monomer at all steps of the reaction pathway and does not transiently dimerize or use sequential transesterification reactions to cleave substrate. Our results are consistent with the I-BmoI DNA-binding domain acting as a molecular anchor to tether the GIY-YIG domain to substrate, permitting rotation of the GIY-YIG domain to sequentially nick each DNA strand. These data highlight the mechanistic differences between monomeric GIY-HEs and dimeric or tetrameric GIY-YIG restriction enzymes, and they have implications for the use of the GIY-YIG domain in genome-editing applications.

Protein function prediction using text-based features extracted from the biomedical literature: the CAFA challenge

BACKGROUND

Advances in sequencing technology over the past decade have resulted in an abundance of sequenced proteins whose function is yet unknown. As such, computational systems that can automatically predict and annotate protein function are in demand. Most computational systems use features derived from protein sequence or protein structure to predict function. In an earlier work, we demonstrated the utility of biomedical literature as a source of text features for predicting protein subcellular location. We have also shown that the combination of text-based and sequence-based prediction improves the performance of location predictors. Following up on this work, for the Critical Assessment of Function Annotations (CAFA) Challenge, we developed a text-based system that aims to predict molecular function and biological process (using Gene Ontology terms) for unannotated proteins. In this paper, we present the preliminary work and evaluation that we performed for our system, as part of the CAFA challenge.

RESULTS

We have developed a preliminary system that represents proteins using text-based features and predicts protein function using a k-nearest neighbour classifier (Text-KNN). We selected text features for our classifier by extracting key terms from biomedical abstracts based on their statistical properties. The system was trained and tested using 5-fold cross-validation over a dataset of 36,536 proteins. System performance was measured using the standard measures of precision, recall, F-measure and overall accuracy. The performance of our system was compared to two baseline classifiers: one that assigns function based solely on the prior distribution of protein function (Base-Prior) and one that assigns function based on sequence similarity (Base-Seq). The overall prediction accuracy of Text-KNN, Base-Prior, and Base-Seq for molecular function classes are 62%, 43%, and 58% while the overall accuracy for biological process classes are 17%, 11%, and 28% respectively. Results obtained as part of the CAFA evaluation itself on the CAFA dataset are reported as well.

CONCLUSIONS

Our evaluation shows that the text-based classifier consistently outperforms the baseline classifier that is based on prior distribution, and typically has comparable performance to the baseline classifier that uses sequence similarity. Moreover, the results suggest that combining text features with other types of features can potentially lead to improved prediction performance. The preliminary results also suggest that while our text-based classifier can be used to predict both molecular function and biological process in which a protein is involved, the classifier performs significantly better for predicting molecular function than for predicting biological process. A similar trend was observed for other classifiers participating in the CAFA challenge.

Programmable in vivo selection of arbitrary DNA sequences

The extraordinary fidelity, sensory and regulatory capacity of natural intracellular machinery is generally confined to their endogenous environment. Nevertheless, synthetic bio-molecular components have been engineered to interface with the cellular transcription, splicing and translation machinery in vivo by embedding functional features such as promoters, introns and ribosome binding sites, respectively, into their design. Tapping and directing the power of intracellular molecular processing towards synthetic bio-molecular inputs is potentially a powerful approach, albeit limited by our ability to streamline the interface of synthetic components with the intracellular machinery in vivo. Here we show how a library of synthetic DNA devices, each bearing an input DNA sequence and a logical selection module, can be designed to direct its own probing and processing by interfacing with the bacterial DNA mismatch repair (MMR) system in vivo and selecting for the most abundant variant, regardless of its function. The device provides proof of concept for programmable, function-independent DNA selection in vivo and provides a unique example of a logical-functional interface of an engineered synthetic component with a complex endogenous cellular system. Further research into the design, construction and operation of synthetic devices in vivo may lead to other functional devices that interface with other complex cellular processes for both research and applied purposes.

Assignment of protein sequences to existing domain and family classification systems: Pfam and the PDB

MOTIVATION

Automating the assignment of existing domain and protein family classifications to new sets of sequences is an important task. Current methods often miss assignments because remote relationships fail to achieve statistical significance. Some assignments are not as long as the actual domain definitions because local alignment methods often cut alignments short. Long insertions in query sequences often erroneously result in two copies of the domain assigned to the query. Divergent repeat sequences in proteins are often missed.

RESULTS

We have developed a multilevel procedure to produce nearly complete assignments of protein families of an existing classification system to a large set of sequences. We apply this to the task of assigning Pfam domains to sequences and structures in the Protein Data Bank (PDB). We found that HHsearch alignments frequently scored more remotely related Pfams in Pfam clans higher than closely related Pfams, thus, leading to erroneous assignment at the Pfam family level. A greedy algorithm allowing for partial overlaps was, thus, applied first to sequence/HMM alignments, then HMM-HMM alignments and then structure alignments, taking care to join partial alignments split by large insertions into single-domain assignments. Additional assignment of repeat Pfams with weaker E-values was allowed after stronger assignments of the repeat HMM. Our database of assignments, presented in a database called PDBfam, contains Pfams for 99.4% of chains >50 residues.

AVAILABILITY

The Pfam assignment data in PDBfam are available at http://dunbrack2.fccc.edu/ProtCid/PDBfam, which can be searched by PDB codes and Pfam identifiers. They will be updated regularly.