Academic literature on the topic 'Pocketome'

The field of targeted protein degradation, through the control of the ubiquitin–proteasome system (UPS), is progressing considerably; to exploit this new therapeutic modality, the proteolysis targeting chimera (PROTAC) technology was born. The opportunity to use PROTACs engaging of new E3 ligases that can hijack and control the UPS system could greatly extend the applicability of degrading molecules. To this end, here we show a potential application of the ELIOT (E3 LIgase pocketOme navigaTor) platform, previously published by this group, for a scaffold-repurposing strategy to identify new ligands for a novel E3 ligase, such as TRIM33. Starting from ELIOT, a case study of the cross-relationship using GRID Molecular Interaction Field (MIF) similarities between TRIM24 and TRIM33 binding sites was selected. Based on the assumption that similar pockets could bind similar ligands and considering that TRIM24 has 12 known co-crystalised ligands, we applied a scaffold-repurposing strategy for the identification of TRIM33 ligands exploiting the scaffold of TRIM24 ligands. We performed a deeper computational analysis to identify pocket similarities and differences, followed by docking and water analysis; selected ligands were synthesised and subsequently tested against TRIM33 via HTRF binding assay, and we obtained the first-ever X-ray crystallographic complexes of TRIM33α with three of the selected compounds.

El trabajo presentado en esta tesis cubre varios campos de investigación relacionados con el desarrollo de moléculas bioactivas. Se compone de cinco partes distintas que se resumen aquí. Predicción de la utilidad farmacológica de dianas terapéuticas. El desarrollo de fármacos está generalmente dirigido a inhibir la función de una proteína específica. Pero para validar esta proteína como diana terapéutica, al principio de un proyecto de descubrimiento de fármacos se tiene que saber si una molécula de tipo fármaco puede unirse con suficiente afinidad a la proteína como para alterar su función. Existen métodos que predicen si una potencial diana terapéutica es tratable o no por vía farmacológica, lo que se ha dado en llamar ‘druggability’. El problema es que estos métodos no están accesibles libremente y su validación es discutible. En la primera parte de la tesis se ha compilado un conjunto extensivo de datos de cavidades en proteínas cuyo novel de ‘druggability’ es conocido, haciéndolo accesible en una plataforma web pública (http://fpocket.sourceforge.net/dcd). Estos datos pueden ser modificados por cualquier persona que quiera contribuir al desarrollo de este conjunto de datos, aumentando su volumen o mejorando su calidad. En estudios previos, los sitios druggable se han asociado a cavidades profundas e hidrofóbicas, ignorando la importancia que tiene los grupos polares en el sitio de unión y su posible relación con la ‘druggability’. Utilizando el set de datos compilado previamente, hemos encontrado que aunque las cavidades ‘druggables’ son mas hidrofóbicas, también tienen grupos polares más expuestos pero con poca superficie de interacción. Esta observación es objeto de posteriores investigaciones en la segunda parte de la tesis. Finalmente, se ha utilizado un algoritmo de búsqueda de sitios de unión, fpocket, que empecé a desarrollar como proyecto de master. Este programa se ha utilizado para extraer todas las características de las cavidades ‘druggables’ y no ‘druggables’ y estos parámetros se han utilizado para entrenar un modelo logístico capaz de predecir si un sitio es druggable o no. Demostramos que el algoritmo y la función de puntuación desarrollado durante esta primera parte predice la ‘druggability’ de manera fiable. Los resultados son de igual calidad a los obtenidos con el único otro programa accesible con funcionalidad parecida (SiteFinder, de Schrödinger), pero nuestro programa tiene las siguientes importantes ventajas: 1) es libre; 2) es mucho más eficiente computacionalmente; y 3) trabaja sobre cavidades detectadas automáticamente por el programa, lo que permite aplicaciones a gran escala. En otras partes de la tesis se verá como su aplicación al conjunto del PDB permite novedosas aplicaciones en el área del diseño de fármacos. Análisis de movilidad de cavidades de las proteínas Existen una gran variedad de algoritmos que permiten identificar posibles sitios de unión en las estructuras tridimensionales de las proteínas. El trabajo presentado en la sección anterior de esta tesis permitía extender uno de estos algoritmos para caracterizar la ‘druggability’ de las cavidades. Un problema de gran calado en el estado actual de la técnica, tanto de detección de sitios de unión como en diseño de fármacos en general,4 es que las proteínas se tratan como un cuerpo rígido a pesar de que en realidad gozan de una gran movilidad estructural. El objetivo de esta sección era otorgar todavía otra funcionalidad a fpocket, el programa de predicción de sitios de unión, para permitir también la detección y el análisis de cavidades de proteínas en movimiento. Habitualmente, los movimientos de las proteína se pueden simular usando la dinámica molecular (MD). Una herramienta capaz de analizar conjuntos de estructuras derivados de MD u otras fuentes puede ser, por tanto, extremadamente útil para observar la aparición de cavidades transitorias y su plasticidad. Como resultado final de este trabajo, se presenta un nuevo programa informático, llamado MDpocket y que se enmarca dentro del paquete fpocket. Para cada conformación de la proteína, se ejecuta un ciclo de detección de cavidades con fpocket. Los resultados de este proceso se plasman sobre una malla tridimensional superpuesta a la estructura de la proteína. La malla puede entonces ser visualizada o analizada en mayor detalle. Lo primero se puede llevar a cabo con programas de visualización molecular tales como PyMOL, VMD o Chimera. Otra funcionalidad dentro de MDpocket es la de seguir la evolución de las propiedades de una cavidad o zona de interés (definida por el usuario) a lo largo del tiempo. Cabe destacar que MDpocket es igualmente capaz de identificar sitios de unión de moléculas tipo fármaco como pequeños canales en la matriz proteica que pueden ser importantes para la migración de pequeños ligandos como gases o moléculas de agua. Las posibles aplicaciones de MDpocket se ejemplifican en tres casos distintos. El primero es la capacidad de identificar la apertura transitoria de cavidades en el sitio de unión de ATP en la proteína HSP90. En el segundo ejemplo, se muestra como MDpocket permite identificar un canal de migración de moléculas biatómicas en mioglobina, un sistema de referencia bien conocido. Aquí se demuestra que MDpocket puede, no tan solo identificar los sitios internos de unión a Xenón, sino también los canales que se abren de forma transitoria para permitir a los ligandos migrar de un sitio a otro. En el último ejemplo, las propiedades del sitio de unión a ATP de la proteína cinasa P38 se analizaron a lo largo de una trayectoria de MD, evaluando la capacidad de MDpocket para identificar aquellas conformaciones que pueden ser particularmente útiles para realizar docking molecular. Uno de los principales problemas en docking de proteína-ligando es que el receptor generalmente se considera rígido, mientras que si se utilizan múltiples conformaciones (cristalográficas o derivadas de MD) para representar al receptor es difícil decidir a priori cuales de ellas pueden dar mejores resultados. Aquí mostramos que MDpocket se puede utilizar para seleccionar conformaciones concretas de una trayectoria de MD para usarlas en procesos de docking. Concretamente, hemos observado que la densidad hidrofóbica promedio (previamente identificada como un descriptor importante para predecir ‘druggability’) correlaciona bien con la probabilidad de que el modo de unión de ligandos pueda ser predicho correctamente. Tal como en el trabajo anterior, MDpocket se incluye dentro del proyecto fpocket y se está accesible como una herramienta libre y de código abierto. Relaciones estructura-cinética de unión El control de los tiempos en interacciones moleculares es una propiedad esencial de los sistemas bioquímicos, pero poco se conoce sobre los factores estructurales que gobiernan la cinética de los procesos de reconocimiento molecular. Partiendo de una observación realizada durante el trabajo de predicción de ‘druggability’, aquí se ha investigado el papel que átomos con poca superficie expuesta a solvente pueden jugar en los sitios de unión de la proteína. En particular, encontramos que los átomos polares en los sitios druggable son minoritarios en comparación con los átomos apolares, pero si bien pueden tener poca superficie accesible, tienden a ser mas protuberantes, lo que los hace más accesibles para establecer interacciones. Hemos establecido que esta propiedad puede estar relacionada con la cinética de unión/disociación de un ligando a su cavidad en el receptor. En diseño de fármacos, la vida media del complejo formado entre el fármaco y su diana terapéutica determina en gran medida sus efectos biológicos, pero en ausencia de relaciones estructura cinética, se hace imposible optimizar esta propiedad de forma racional. Aquí se muestra que átomos polares prácticamente enterrados (ABPAs) – un elemento comúnmente encontrado en los sitios de unión de proteínas – tienden a formar puentes de hidrógeno que están protegidos de las moléculas de agua. La formación y ruptura de este tipo de puentes de hidrógeno implica un estado de transición penalizado energéticamente porque ocurre de modo asincrónico con el proceso de deshidratación/rehidratación. En consecuencia, los puentes de hidrógeno protegidos se intercambian a velocidades lentas. Estas conclusiones se basan en el estudio computacional del proceso de unión de un pequeño ligando a un sitio de unión modelo. El receptor modelo se construyó para permitir modular tanto el grado de exposición del átomo polar como la curvatura del entorno apolar. Mediante el uso de dinámicas moleculares con constricciones y la relación de Jarzinsky, se obtuvieron los perfiles de energía libre de unión para cada cavidad. La presencia de un estado de transición (y por tanto menor velocidad de asociación/disociación) puede anticiparse mediante un simple análisis estructural tal como la medición de la superficie accesible del átomo polar o su grado de protrusión. Esto constituye una nueva y valiosa clave para interpretar y predecir relaciones estructura-actividad, que se ha puesto a prueba investigando sistemas reales. En primer lugar, analizando tanto estructuras cristalográficas depositadas en el PDB como trayectorias de dinámica molecular, se ha demostrado que aquellas moléculas de agua que forman puentes de hidrógeno con ABPAs tienden a tener menor movilidad e intercambios más lentos. Posteriormente, la validez del principio se ha demostrado en dos pares de inhibidores de la proteína Hsp90, una diana terapéutica para cáncer, para los que se han obtenido datos estructurales, termodinámicos y cinéticos mediante distintas técnicas experimentales. El acuerdo entre observables macroscópicas y los resultados de simulaciones moleculares confirma la función de los puentes de hidrógeno protegidos de solvente como trampas cinéticas e ilustra como nuestro hallazgo puede ser usado para facilitar el proceso de diseño de fármacos basado en estructura. Base de datos de cavidades: hacia el ‘pocketoma’ El trabajo presentado en el principio de la tesis perseguía un objetivo muy especifico, que se enmarca en un proyecto mayor del grupo de investigación. La herramienta de predicción de ‘druggability’ se desarrolló con el fin de cribar grandes bases de datos estructurales, tales como el PDB, identificando complejos proteína-proteína no obligados (transitorios) que contengan en su interfase una cavidad potencialmente capaz de unir moléculas de tipo fármaco. Con ello se pretende estabilizar selectivamente dicha interacción y conseguir un efecto biológico que pueda ser terapéutico. Dado que la información sobre cavidades y su druggability asociada va a ser explotadas por otras personas en el grupo en este y otro tipo de proyectos encaminados a facilitar la explotación de nuevos mecanismos de acción, es necesario crear una base de datos que contenga esta información y sea fácilmente navegable. Para empezar, se ejecutó el programa para cada una de las estructuras depositadas en el PDB, identificando todas las cavidades y extrayendo sus descriptores, entre los que se incluye la función de puntuación de druggability. Esta información se guarda de forma organizada en una base de datos relacional (pocketDB), que se relaciona con otras bases de datos tales como Uniprot y Uniref, que contienen información sobre secuencias. Igualmente, se incluyeron información sobre la estructura cuaternaria y otros recursos tales como Kegg, haciendo de pocketDB un potente recurso para filtrar de modo eficiente millones de cavidades, identificando aquellas que tengan mayor interés para cada proyecto. De modo particular, cabe destacar la aplicación de pocketDB a la identificación de cavidades ‘druggable’ situadas en la interfase de complejos transitorios proteína-proteína, resultando en 39 complejos candidatos, entre los que se recobraron 3 casos conocidos previamente, lo que valida la metodología. Además otros tres sistemas identificados, después de una inspección minuciosa, han sido seleccionados en el grupo como candidatos para realizar la prueba de concepto que valide esta nueva estrategia farmacológica. Uno de ellos está actualmente en fase de validación experimental. El ‘pocketoma’ La última sección de mi tesis presenta un proyecto que hace un uso extensivo de la base de datos de cavidades (pocketDB) previamente presentada. Dos cuestiones importantes aún persisten hoy día en el proceso de descubrimiento de fármacos y, de algún modo, contribuyen a la alta tasa de fracasos en las fases tardías del desarrollo, que normalmente se explican por la baja eficacia o el exceso de efectos secundarios (normalmente tóxicos) para el organismo. Ambas situaciones pueden ser explicadas por un fenómeno común: la falta de selectividad. Las fármacos interaccionan en la célula con una diversa variedad de macromoléculas además de con la diana terapéutica, induciendo así efectos secundarios imprevistos. Asimismo, considerar solamente una diana para nuestro fármaco puede resultar menos efectivo de lo esperado, puesto que la pérdida de función de una sola proteína diana puede ser fácilmente reemplazada debido a los mecanismos homeostáticos controlados por robustas redes de interacciones, derivando en una pérdida de eficacia de nuestra molécula. Este trabajo pretende sentar las bases para poder establecer relaciones entre macromoléculas biológicas en base a su potencial para interaccionar con una misma entidad química (fármaco). El trabajo se basa en la asunción de que cavidades similares son capaces de unir ligandos similares. Existen varios métodos que calculan la similitud entre cavidades de unión, sin embargo, hasta la fecha no se ha realizado un análisis relacional profundo de todas las cavidades en el PDB que permita establecer relaciones entre ellos. Con el fin de cumplir con los requerimientos técnicos de unos objetivos tan ambiciosos, se ha desarrollado un novedoso método de comparación de cavidades. En este particular método se considera una representación muy abstracta de la cavidad. Dicha representación reúne información tanto sobre la forma de la cavidad cómo sobre la distribución por pares de los puntos de interacción en la superficie. No obstante, la información sobre la topología exacta de la cavidad es ignorada. Tal abstracción intenta relacionar cavidades que estructuralmente están alejadas, pero que se parecen entre ellas en términos de forma y propiedades fisicoquímicas globales. Usando varios ejemplos de validación en un conjunto de cavidades de referencia, el método resulta capaz de recuperar de un gran set de cavidades aquellas más similares y relacionadas entre sí. Del mismo modo, el método demuestra ser útil para encontrar relaciones entre cavidades previamente no relacionadas y que sin embargo unen los mismos ligandos o moléculas muy similares. Esta nueva implementación se ha utilizado en un experimento de exploración a gran escala para encontrar (i) las mismas cavidades en diferentes estructuras, (ii) cavidades relacionadas y (iii) cavidades con la misma estructura de ligando. Se obtuvieron excelentes resultados para estas tres categorías. Seguidamente, se compararon entre ellas todas las cavidades encontradas en el PDB. Se desarrolló una herramienta computacional novedosa para permitir la navegación en el 'pocketoma' resultante de las comparaciones, que será posible descargar gratuitamente. Los resultados presentados muestran por primera vez un espacio global interrelacionando cavidad-ligando para todas las cavidades del PDB. Finalmente, se señalan a continuación dos aplicaciones que ponen de manifiesto el posible impacto del ‘pocketoma’ y la herramienta de navegación. En el primer ejemplo, una cavidad no caracterizada encontrada en GSK3-β fue comparada contra todas las cavidades del PDB, permitiendo obtener una variedad de cavidades, y sistemas, supuestamente relacionadas. Entre los resultados se encontraban las subunidades de unión a ADP dhaL y dhaM de la PTS dependiente di-hidroacetona cinasa. Se encontró que el sitio de unión a ADP era muy similar a la cavidad investigada en GSK3-β con una sorprendente similitud estructural entre ambos. En el último ejemplo, se navegó por el ‘pocketoma’ utilizando la herramienta visual desarrollada para tal tarea. Durante la exploración se encontró una curiosa e importante relación entre el sitio de unión de la hormona del receptor de estrógenos y una cavidad no caracterizada en la proteína Caspasa-3. Seguidamente, se realizó un estudio de docking con ligandos similares a la hormona y se procedió a realizar una extensiva dinámica molecular con el ligando en la mejor posición para verificar la estabilidad del compuesto en la cavidad de Caspasa-3. El complejo proteína-ligando estudiado resultó ser muy estable. Actualmente, el compuesto identificado se encuentra bajo validación experimental por nuestros colaboradores.

BIOLOGICAL processes are governed through specific interactions of macromolecules. The three-dimensional structural information of the macromolecules is necessary to understand the basis of molecular recognition. A large number of protein structures have been determined at a high resolution using various experimental techniques such as X-ray crystallography, NMR, electron microscopy and made publicly available through the Protein Data Bank. In the recent years, comprehending function by studying a large number of related proteins is proving to be very fruitful for understanding their biological role and gaining mechanistic insights into molecular recognition. Availability of large-scale structural data has indeed made this task of predicting the protein function from three-dimensional structure, feasible. Structural bioinformatics, a branch of bioinformatics, has evolved into a separate discipline to rationalize and classify the information present in three-dimensional structures and derive meaningful biological insights. This has provided a better understanding of biological processes at a higher resolution in several cases. Most of the structural bioinformatics approaches so far, have focused on fold-level analysis of proteins and their relationship to sequences. It has long been recognized that sequence-fold or fold-function relationships are highly complex. Information on one aspect cannot be readily extrapolated to the other. To a significant extent, this can be overcome by understanding similarities in proteins by comparing their binding site structures. In this thesis, the primary focus is on analyzing the small-molecule ligand binding sites in protein structures, as most of the biological processes ranging from enzyme catalysis to complex signaling cascades are mediated through protein-ligand interactions. Moreover, given that the precise geometry and the chemical properties of the residues at the ligand binding sites dictate the molecular recognition capabilities, focusing on these sites at the structural level, is likely to yield more direct insights on protein function. The study of binding sites at the structural level poses several problems mainly because the residues at the site may be sequentially discontinuous but spatially proximal. Further, the order of the binding site residues in primary sequence, in most of cases has no significance for ligand binding. Compounding these difficulties are additional factors such as, non-uniform contribution to binding from different residues, and size-variations in binding sites even across closely related proteins. As a result, methods available to study ligand-binding sites in proteins, especially on a large-scale are limited, warranting exploration of new approaches. In the present work, new methods and tools have been developed to address some of these challenges in binding site analysis. First, a novel tool for site-based function annotation of protein structures, called PocketAnnotate was developed ( http://proline.biochem.iisc.ernet. in/pocketannotate/). PocketAnnotate, detects the putative binding sites from a given protein structure and compares them to known binding sites in PDB to derive functional annotation in terms of ligand association. Since the tool derives functional annotation at the level of binding sites, it has an advantage over other methods that solely utilize fold or sequence information. This becomes even more important for cases where there is no detectable homology with entries in existing databases, as Pocket Annotate does not depend on evolutionary based information for annotation. Second, a web-accessible tool for in silico almandine scanning mutations of binding site residues called ABS-Scan has been developed ( http://proline.biochem.iisc.ernet.in/abscan/). This tool helps in assessing the contribution of the individual residues of binding sites in the protein towards ligand recognition. All residues, one at a time, in a binding site are mutated systematically to an alanine and the ability of the corresponding mutant to bind a given ligand is analyzed. The contribution of each residue towards ligand binding is calculated through a G value derived by comparing the binding affinity to the wild-type protein-ligand complex. Third, a database called Protein-Ligand Interaction Clusters (PLIC) has been developed to identify and analyze the information of similarity across binding sites in PDB, which has been provided in the form of a web-accessible database ( http://proline.biochem.iisc.ernet/ PLIC). Protein-ligand interactions are primarily explored using three different computational approaches - (i) binding site characteristics including pocket shape, nature of residues and interaction profiles with different kinds of chemical probes, (ii) atomic contacts between protein and ligands (iii) binding energetics involved in interactions derived from scoring functions developed for docking. The information on variations in these features derived from different computational tools is also included in the database for enabling the characterization of the binding sites. As a case study to demonstrate the usefulness of these tools, they have been applied to decipher the complexity of S-adenosyl methionine interactions with the protein. Around 1,213 binding sites of SAM or SAM-like compounds could be extracted from the PLIC database. The SAM or SAM-like compounds were observed to interact with ∼18 different protein-fold types. The variations in different protein-ligand contacts across fold types were analyzed. The fold-specific interaction properties and contribution of individual residues towards SAM binding are identified. The tools developed and example analyses using them are described in Chapter 2. Chapter 3 describes a large-scale pocketome analysis from structural complexes in PDB, in an effort to characterize the known pocket space of protein-ligand interactions. Tools devel-opted as described in Chapter 2 are used for this. A set of 84,846 binding sites compiled from PDB, have been comprehensively analyzed with an objective of obtaining (a) classification of binding sites, (b) sequence-fold-site relationships among proteins, (c) a minimal set of physicochemical attributes sufficient to explain ligand recognition specificity and (d) site-type specific signatures in terms of physicochemical features. A new method to describe binding sites was developed in the form of BScIds such that the structural fold information is well captured. Binding sites and similarities among them were abstracted in the form of networks where each node represents a binding site and an edge between two nodes represents significant similarity between the sites at the structural level. Pocketome networks were constructed from the large-scale information on protein-ligand interactions in the PLIC database. The large pocketome network was then studied to derive relationships between protein folds and chemical entities they interact with. A classification of the binding pockets was achieved by analyzing the pocketome network using graph theoretical approaches combined with clustering methods. 10,858 clusters were identified from the network, each indicating a site-type. Thus, it can be said that there are about 10,858 site-types. Classification of ligand associations into specific site-types helps greatly in resolving the complex relationships by yielding specific site-type ligand associations. The observed classification was further probed to understand the basis of ligand recognition by representing the pockets through feature vectors. These features capture a wide range of physicochemical properties that can be used to derive site-type specific signatures and explore the pocket-space of protein-ligand interactions. A principal component analysis of these features reveals that binding site feature space is continuous in the entire PDB and minor changes in specific features can give rise to significant differences in ligand specificity, consequently defining their distinct functional roles. The weights were also derived for these features through the use of different information theoretic approaches to explain the multiple-specificity of protein-ligand interactions. Analysis of binding sites arising from contribution of residues from different protein fold-types revealed increasing diversity of physicochemical properties at the site, supporting the hypothesis that combination of folds could give rise to new binding sites. Given that a finer appreciation of the molecular mechanisms within the cell is possible only with the structural information, the next objective was to explore if a structural view of an entire proteome can be obtained and if a pocketome could be constructed and analyzed. With this in mind, the causative agent of tuberculosis - Mycobacterium tuberculosis (Mtb) was chosen. Mtb is also being studied in the laboratory from a systems biology perspective, which enabled exploration of how systems and the structural perspectives could be combined and applied for drug discovery. Chapters 4 to 6 describe this effort. The genome sequence of Mycobacterium tuberculosis (Mtb) H37Rv, indicates the presence of ∼4,000 protein coding genes, of which experimentally determined structures are available for ∼300 proteins. Further, advances in homology modeling methods have made it feasible to obtain structural models for many more proteins in the proteome. Chapter 4 describes the efforts for obtaining the Mtb structural proteome, through which the three-dimensional struc-tures were derived for ∼70% of the proteins in the genome. Functional annotation of each protein was derived based on fold-based functional assignments, binding-site comparisons and consequent ligand associations. PocketAnnotate, a site-based function annotation pipeline was utilized for this purpose and is described in Chapter 2. Besides these, the annotation covers detection of various sequence and sub-structural motifs and quaternary structure predictions based on the corresponding templates. The study provides a unique opportunity to obtain a global perspective of the fold distribution in the genome. The annotation indicates that cellular metabolism can be achieved with only 219 unique folds. New insights about the folds that predominate in the genome, as well as the fold-combinations that make up multi-domain proteins are also obtained. 1,728 binding pockets have been associated with ligands through binding site identification and sub-structure similarity analyses, yielding a list of ligands that can participate in various biochemical events in the mycobacterial cell. A web-accessible database MtbStructuralproteome has been developed to make the data and the analyses available to the community, ( http://proline.physics.iisc.ernet.in/Tbstructuralannotation). The resource, being one of the first to be based on structure-based functional annotations at a genome scale, is expected to be useful for better understanding of tuberculosis and for application in drug discovery. The reported annotation pipeline is fairly generic and can be applied to other genomes as well. Chapter 5 describes the characterization of the Mtb pocketome. For the structural models of the Mtb proteome described in chapter 4, a genome-scale binding site prediction exercise was carried out using three different computational methods and subsequently obtaining consensus predictions. The three methods were independent and were based on considering geometry, inter-molecular energies with probes and sequence conservations in evolutionarily related proteins respectively. In all, 13,858 consensus binding pockets were predicted in 2,877 proteins. The pocket space within Mtb was then explored through systematic all-pair comparisons of binding sites. The number of site-types within Mtb was found to be 6,584, as compared to the ∼400 structural folds and 1,831 unique sequence families. This reveals that the pocket space is larger than the sequence or fold-space, suggesting that variations at the site-level contribute significantly to functional repertoire of the organism. By comparing the pockets with the PDB sites enclosing known ligands, around 6906 binding sites were observed to exhibit significant similarity in the entire pockets to some or the other known binding site in PDB. 1,213 metabolites could be mapped onto 665 enzymes covering most of the metabolic pathways. The identified ligands serve as a predicted metabolome for unit abundances of the proteins. A list of proteins containing unique pockets is also identified. The binding pockets, similarities they share within Mtb and the ligands mapped onto them are all made available in a web-accessible database at http://proline.biochem.iisc.ernet.in/mtbpocketome/. The availability of structural information of the pocketome at a genome-scale opens up several opportunities in drug discovery. They can be directly applied for understanding mechanism of drug action, predicting adverse effects and pharmacodynamics of a drug. Moreover, it enables exploration of new ideas in drug discovery. Polypharmacology is a new concept that aims at modulating multiple drug targets through a single chemical entity. Currently, there are no established approaches to either select appropriate target sets or design polypharmacological drugs. In this study, a structural-proteomics approach is explored to first characterize the pocketome and then utilize it to identify similar binding sites. The knowledge of similarity relationships between the binding sites within the genome can be used in identifying possible polypharmacological drug targets. A pocket similarity based clustering of binding site residues resulted in identification of binding site sets, each having a theoretical potential to interact with a common ligand. A polypharmacological index was formulated to rank targets by incorporating a measure of drug ability and similarity to other pockets within the proteome. By comparing with known drug binding sites from databases such as the Drug Bank, the study has yielded a ready shortlist that includes sets of promising drug targets with polypharmacological possibilities and at the same time has identified possible drug candidates either directly for repurposing or at the least as significant lead clues that can be used to design new drug molecules against the entire group of proteins in each set. This analysis presents a rational approach to identify targets with polypharmacological potential, clues about lead compounds and a list of candidates for drug repurposing. This thesis demonstrates the feasibility of utilizing the structural bioinformatics approaches at a genome-scale. The tools developed for analyzing large-scale data on protein-ligand inter-actions could be applied to characterize the pocket-space of protein-ligand interactions. The network theory approaches applied in this work, make large-scale data tractable and enable binding-site typing. The binding site analysis at a genome-scale for Mtb is first of its kind and has provided novel insights into the pocket space. The binding site analysis performed on a genome-scale for Mtb provided an opportunity to rationalize the polypharmacological target selection and explore drugs for repurposing in TB. In the larger context, structural modelling of a proteome, mapping the small-molecule binding space in it and understanding the determinants of small-molecule recognition forms a major step in defining a proteome at higher resolution. This in turn will serve as a valuable input towards the emerging field of structural-systems biology, which seeks to understand the biological models at a systems level without compromising on the resolution of the study.

BIOLOGICAL processes are governed through specific interactions of macromolecules. The three-dimensional structural information of the macromolecules is necessary to understand the basis of molecular recognition. A large number of protein structures have been determined at a high resolution using various experimental techniques such as X-ray crystallography, NMR, electron microscopy and made publicly available through the Protein Data Bank. In the recent years, comprehending function by studying a large number of related proteins is proving to be very fruitful for understanding their biological role and gaining mechanistic insights into molecular recognition. Availability of large-scale structural data has indeed made this task of predicting the protein function from three-dimensional structure, feasible. Structural bioinformatics, a branch of bioinformatics, has evolved into a separate discipline to rationalize and classify the information present in three-dimensional structures and derive meaningful biological insights. This has provided a better understanding of biological processes at a higher resolution in several cases. Most of the structural bioinformatics approaches so far, have focused on fold-level analysis of proteins and their relationship to sequences. It has long been recognized that sequence-fold or fold-function relationships are highly complex. Information on one aspect cannot be readily extrapolated to the other. To a significant extent, this can be overcome by understanding similarities in proteins by comparing their binding site structures. In this thesis, the primary focus is on analyzing the small-molecule ligand binding sites in protein structures, as most of the biological processes ranging from enzyme catalysis to complex signaling cascades are mediated through protein-ligand interactions. Moreover, given that the precise geometry and the chemical properties of the residues at the ligand binding sites dictate the molecular recognition capabilities, focusing on these sites at the structural level, is likely to yield more direct insights on protein function. The study of binding sites at the structural level poses several problems mainly because the residues at the site may be sequentially discontinuous but spatially proximal. Further, the order of the binding site residues in primary sequence, in most of cases has no significance for ligand binding. Compounding these difficulties are additional factors such as, non-uniform contribution to binding from different residues, and size-variations in binding sites even across closely related proteins. As a result, methods available to study ligand-binding sites in proteins, especially on a large-scale are limited, warranting exploration of new approaches. In the present work, new methods and tools have been developed to address some of these challenges in binding site analysis. First, a novel tool for site-based function annotation of protein structures, called PocketAnnotate was developed ( http://proline.biochem.iisc.ernet. in/pocketannotate/). PocketAnnotate, detects the putative binding sites from a given protein structure and compares them to known binding sites in PDB to derive functional annotation in terms of ligand association. Since the tool derives functional annotation at the level of binding sites, it has an advantage over other methods that solely utilize fold or sequence information. This becomes even more important for cases where there is no detectable homology with entries in existing databases, as Pocket Annotate does not depend on evolutionary based information for annotation. Second, a web-accessible tool for in silico almandine scanning mutations of binding site residues called ABS-Scan has been developed ( http://proline.biochem.iisc.ernet.in/abscan/). This tool helps in assessing the contribution of the individual residues of binding sites in the protein towards ligand recognition. All residues, one at a time, in a binding site are mutated systematically to an alanine and the ability of the corresponding mutant to bind a given ligand is analyzed. The contribution of each residue towards ligand binding is calculated through a G value derived by comparing the binding affinity to the wild-type protein-ligand complex. Third, a database called Protein-Ligand Interaction Clusters (PLIC) has been developed to identify and analyze the information of similarity across binding sites in PDB, which has been provided in the form of a web-accessible database ( http://proline.biochem.iisc.ernet/ PLIC). Protein-ligand interactions are primarily explored using three different computational approaches - (i) binding site characteristics including pocket shape, nature of residues and interaction profiles with different kinds of chemical probes, (ii) atomic contacts between protein and ligands (iii) binding energetics involved in interactions derived from scoring functions developed for docking. The information on variations in these features derived from different computational tools is also included in the database for enabling the characterization of the binding sites. As a case study to demonstrate the usefulness of these tools, they have been applied to decipher the complexity of S-adenosyl methionine interactions with the protein. Around 1,213 binding sites of SAM or SAM-like compounds could be extracted from the PLIC database. The SAM or SAM-like compounds were observed to interact with ∼18 different protein-fold types. The variations in different protein-ligand contacts across fold types were analyzed. The fold-specific interaction properties and contribution of individual residues towards SAM binding are identified. The tools developed and example analyses using them are described in Chapter 2. Chapter 3 describes a large-scale pocketome analysis from structural complexes in PDB, in an effort to characterize the known pocket space of protein-ligand interactions. Tools devel-opted as described in Chapter 2 are used for this. A set of 84,846 binding sites compiled from PDB, have been comprehensively analyzed with an objective of obtaining (a) classification of binding sites, (b) sequence-fold-site relationships among proteins, (c) a minimal set of physicochemical attributes sufficient to explain ligand recognition specificity and (d) site-type specific signatures in terms of physicochemical features. A new method to describe binding sites was developed in the form of BScIds such that the structural fold information is well captured. Binding sites and similarities among them were abstracted in the form of networks where each node represents a binding site and an edge between two nodes represents significant similarity between the sites at the structural level. Pocketome networks were constructed from the large-scale information on protein-ligand interactions in the PLIC database. The large pocketome network was then studied to derive relationships between protein folds and chemical entities they interact with. A classification of the binding pockets was achieved by analyzing the pocketome network using graph theoretical approaches combined with clustering methods. 10,858 clusters were identified from the network, each indicating a site-type. Thus, it can be said that there are about 10,858 site-types. Classification of ligand associations into specific site-types helps greatly in resolving the complex relationships by yielding specific site-type ligand associations. The observed classification was further probed to understand the basis of ligand recognition by representing the pockets through feature vectors. These features capture a wide range of physicochemical properties that can be used to derive site-type specific signatures and explore the pocket-space of protein-ligand interactions. A principal component analysis of these features reveals that binding site feature space is continuous in the entire PDB and minor changes in specific features can give rise to significant differences in ligand specificity, consequently defining their distinct functional roles. The weights were also derived for these features through the use of different information theoretic approaches to explain the multiple-specificity of protein-ligand interactions. Analysis of binding sites arising from contribution of residues from different protein fold-types revealed increasing diversity of physicochemical properties at the site, supporting the hypothesis that combination of folds could give rise to new binding sites. Given that a finer appreciation of the molecular mechanisms within the cell is possible only with the structural information, the next objective was to explore if a structural view of an entire proteome can be obtained and if a pocketome could be constructed and analyzed. With this in mind, the causative agent of tuberculosis - Mycobacterium tuberculosis (Mtb) was chosen. Mtb is also being studied in the laboratory from a systems biology perspective, which enabled exploration of how systems and the structural perspectives could be combined and applied for drug discovery. Chapters 4 to 6 describe this effort. The genome sequence of Mycobacterium tuberculosis (Mtb) H37Rv, indicates the presence of ∼4,000 protein coding genes, of which experimentally determined structures are available for ∼300 proteins. Further, advances in homology modeling methods have made it feasible to obtain structural models for many more proteins in the proteome. Chapter 4 describes the efforts for obtaining the Mtb structural proteome, through which the three-dimensional struc-tures were derived for ∼70% of the proteins in the genome. Functional annotation of each protein was derived based on fold-based functional assignments, binding-site comparisons and consequent ligand associations. PocketAnnotate, a site-based function annotation pipeline was utilized for this purpose and is described in Chapter 2. Besides these, the annotation covers detection of various sequence and sub-structural motifs and quaternary structure predictions based on the corresponding templates. The study provides a unique opportunity to obtain a global perspective of the fold distribution in the genome. The annotation indicates that cellular metabolism can be achieved with only 219 unique folds. New insights about the folds that predominate in the genome, as well as the fold-combinations that make up multi-domain proteins are also obtained. 1,728 binding pockets have been associated with ligands through binding site identification and sub-structure similarity analyses, yielding a list of ligands that can participate in various biochemical events in the mycobacterial cell. A web-accessible database MtbStructuralproteome has been developed to make the data and the analyses available to the community, ( http://proline.physics.iisc.ernet.in/Tbstructuralannotation). The resource, being one of the first to be based on structure-based functional annotations at a genome scale, is expected to be useful for better understanding of tuberculosis and for application in drug discovery. The reported annotation pipeline is fairly generic and can be applied to other genomes as well. Chapter 5 describes the characterization of the Mtb pocketome. For the structural models of the Mtb proteome described in chapter 4, a genome-scale binding site prediction exercise was carried out using three different computational methods and subsequently obtaining consensus predictions. The three methods were independent and were based on considering geometry, inter-molecular energies with probes and sequence conservations in evolutionarily related proteins respectively. In all, 13,858 consensus binding pockets were predicted in 2,877 proteins. The pocket space within Mtb was then explored through systematic all-pair comparisons of binding sites. The number of site-types within Mtb was found to be 6,584, as compared to the ∼400 structural folds and 1,831 unique sequence families. This reveals that the pocket space is larger than the sequence or fold-space, suggesting that variations at the site-level contribute significantly to functional repertoire of the organism. By comparing the pockets with the PDB sites enclosing known ligands, around 6906 binding sites were observed to exhibit significant similarity in the entire pockets to some or the other known binding site in PDB. 1,213 metabolites could be mapped onto 665 enzymes covering most of the metabolic pathways. The identified ligands serve as a predicted metabolome for unit abundances of the proteins. A list of proteins containing unique pockets is also identified. The binding pockets, similarities they share within Mtb and the ligands mapped onto them are all made available in a web-accessible database at http://proline.biochem.iisc.ernet.in/mtbpocketome/. The availability of structural information of the pocketome at a genome-scale opens up several opportunities in drug discovery. They can be directly applied for understanding mechanism of drug action, predicting adverse effects and pharmacodynamics of a drug. Moreover, it enables exploration of new ideas in drug discovery. Polypharmacology is a new concept that aims at modulating multiple drug targets through a single chemical entity. Currently, there are no established approaches to either select appropriate target sets or design polypharmacological drugs. In this study, a structural-proteomics approach is explored to first characterize the pocketome and then utilize it to identify similar binding sites. The knowledge of similarity relationships between the binding sites within the genome can be used in identifying possible polypharmacological drug targets. A pocket similarity based clustering of binding site residues resulted in identification of binding site sets, each having a theoretical potential to interact with a common ligand. A polypharmacological index was formulated to rank targets by incorporating a measure of drug ability and similarity to other pockets within the proteome. By comparing with known drug binding sites from databases such as the Drug Bank, the study has yielded a ready shortlist that includes sets of promising drug targets with polypharmacological possibilities and at the same time has identified possible drug candidates either directly for repurposing or at the least as significant lead clues that can be used to design new drug molecules against the entire group of proteins in each set. This analysis presents a rational approach to identify targets with polypharmacological potential, clues about lead compounds and a list of candidates for drug repurposing. This thesis demonstrates the feasibility of utilizing the structural bioinformatics approaches at a genome-scale. The tools developed for analyzing large-scale data on protein-ligand inter-actions could be applied to characterize the pocket-space of protein-ligand interactions. The network theory approaches applied in this work, make large-scale data tractable and enable binding-site typing. The binding site analysis at a genome-scale for Mtb is first of its kind and has provided novel insights into the pocket space. The binding site analysis performed on a genome-scale for Mtb provided an opportunity to rationalize the polypharmacological target selection and explore drugs for repurposing in TB. In the larger context, structural modelling of a proteome, mapping the small-molecule binding space in it and understanding the determinants of small-molecule recognition forms a major step in defining a proteome at higher resolution. This in turn will serve as a valuable input towards the emerging field of structural-systems biology, which seeks to understand the biological models at a systems level without compromising on the resolution of the study.

Biological processes are governed by highly specific macromolecular interactions. Understanding the precise mechanism of ligand recognition in proteins is essential for deriving features responsible for such recognition capabilities. Although protein sequences give first-hand information about their function, their three-dimensional structures, which are the evolutionary units, convey the function better. Three-dimensional structures of many proteins determined through X-ray crystallography and/or NMR are available in the Protein Data Bank, a public repository. This resource has stimulated the development of computational techniques to read and analyse the wealth of structural data. Structural bioinformatics is an area that provides a means to transform information in the protein structures into functional insights and enable addressing a variety of questions about what defines and dictates ligand recognition. Large-scale analyses of several protein-ligand complexes have indicated that both one-to-many and many-to-one relationships of protein-folds and ligand-types are widely seen in the PDB. This means that a given ligand can be recognized by diverse proteins and a given protein can recognize different types of ligands at the same location, ligands referring to endogenous ligands, natural metabolites as well as small molecule inhibitors, and drugs. Given this, it is important to understand the determinants of recognition of a given ligand. This becomes important for applications in drug discovery that includes, lead design and lead optimization, assessment of draggability of a target, identification of off-target effects, polypharmacological targets and drug repurposing. The present work utilizes the information present at the functional sites, rationalizing many examples of ligand binding and deriving useful patterns that can be used for genome-wide function annotation and drug discovery applications. A large-scale analysis of the binding of two important classes of ligands, a sugar and nucleotides was carried out by analysing the sub-structures at their binding sites by matching, aligning and clustering. The two ligands studied are sialic acid, and nucleoside mon/di/tri phosphates (nucleotides or NTPs), for their binding to many proteins reported in the PDB. Sialic acid was found to be recognized by 170 different proteins representing 17 unique sequence families. Our approach deciphered a unified understanding of the basis of recognition of this ligand and showed six structural motifs, which contained different combinations of one or more key structural features, over a common scaffold. The site features refer to certain residues in the binding site that are seen to most frequently occur at their respective topological positions, a result that was evident upon binding site comparisons and 3-D alignment of sites in the different proteins. In the case of nucleotide ligands, 4,677 structures of protein-nucleotide complexes from PDB, belonging to 145 different structural folds and 394 sequence families were analysed, and our results indicated that the sites from diverse proteins group into 27 site-types and further into nine super-types having a structural motif for each site-type. The identified motifs were highly specific when scanned against the entire PDB. A computational alanine mutagenesis study indicated that residues identified to be highly conserved in the motifs also contribute most to binding. Alternate orientations of the ligand in several site-types were observed and rationalized, indicating the possibility of some residues serving as anchors for NTP recognition. Many examples of convergent evolution were identified through this analysis. Next, in order to find the entire set of all binding sites in the proteins that have known 3-D structures, a large-scale analysis of 30457 representative protein structures in PDB was carried out by detecting additional binding sites in them, followed by a site-comparison to recognize unexpected similarities. The pocket-space currently defined in PDB is incomplete, as binding-sites remain uncharacterized in many proteins despite the availability of their structures. To bridge this gap, we computationally detect pockets in the entire PDB and present a comprehensive resource of an ‘augmented pocketome’ consisting of 249,096 pockets (http://proline.biochem.iisc.ernet.in/PocketDB), which shows a 2-fold increase in fold coverage and a 7-fold increase in pocket-space when compared to what is currently known. Possible ligand associations for about 56% of these pockets were deduced. A clustering of the pocketome led to the identification of 2,161 site-types, and the associated ligands into 1,037 ligand-types, which together provided fold-to-site-type-to-ligand-type associations. The resource facilitates a structure-based function annotation, delineation of a structural basis of ligand recognition, and provides functional clues for Domain of Unknown Function (DUFs), allosteric proteins and druggable pockets. Next, it was of interest to study the entire binding site-space in a proteome of a given organism. Mycobacterium tuberculosis (Mtb) is a deadly pathogen being studied from multiple perspectives in the laboratory. Structural models of about 70% of the proteome and a set of 13,858 small molecule binding pockets were already available from a previous study in the laboratory. To systematically identify druggable targets in this category, a genome-wide screen was carried out to comprehensively identify all proteins binding to NTP ligands. Being equipped on one hand, with the knowledge of binding site motifs, and on the other, with the structural models of Mtb proteins at a genome-scale, this exercise was made possible. A total of 1,768 proteins in Mtb (about 43% of the proteome) were predicted to bind NTP ligands, which constitute the NTPome. Using an experimental proteomics approach involving dye-ligand affinity chromatography, 47 different proteins were validated, of which four are hypothetical proteins. This analysis also provides the precise list of binding site residues in each case, and the probable ligand binding pose. As the list includes several known and potential drug targets, the identification of NTP binding can directly facilitate structure-based drug design of these targets. Further, 305 proteins in the NTPome were identified as important drug targets. A successful drug discovery program stands on the first principles of good target identification. Previous studies carried out in the laboratory combined with contributions from this thesis work, a promising target, nucleoside diphosphate kinase (NDK), that possesses many of the desirable properties of an ideal target, was identified. Using the NTP motifs previously derived, the nucleotide pocket in NDK was studied for the most important residues that confer binding, thus prioritizing the pocket residues for mediating crucial interactions with the ligand. Next, a screen of the nucleotide pocket of NDK against the drug-binding sites known in PDB resulted in identifying a handful of potential screening candidates that contained several elements of an ideal ligand for the NDK pocket. A thorough biochemical characterization using protein purification and subsequent quantification by a spectrophotometric coupled assay, led to the experimental testing of the identified candidates against NDK enzyme activity. This resulted in identifying enalapril and cloxacillin as NDK inhibitors. Of the two compounds, enalapril was efficient in 100% inhibition of NDK activity, while cloxacillin showed a maximum inhibition of 78%. A test for inhibition of Mtb H37Rv growth resulted in determining cell inhibition for both compounds, of which cloxacillin was known previously. Thus, by using the principles of structure-based drug design, two lead molecules were identified, which are reported for the first time for Mtb NDK. Thus, using a structural bioinformatics approach, a large-scale characterization of small-molecule binding sites has been carried out in this thesis work. The work starts from identifying ligand recognition principles of proteins and understanding the molecular capabilities of ligand binding in diverse proteins, which forms the crux of protein-ligand interaction profiling. Using the sub-structures for recognizing function in uncharacterized proteins, the work carried out here presents a pipeline for function annotation. One goal of understanding the functional capabilities of macromolecules is to design and develop drugs, which have also been dealt with in this thesis work, housing the principles of target identification, druggability and finally, lead identification

In this chapter, we introduce alternative ways to access digital audio collections. We give an overview of existing applications based on tow-dimensional, map-like representations of music collections. Further, we explain two applications for accessing audio files that are based on the Self-Organising Map, an unsupervised neural network model. These two applications—PlaySOM and PocketSOM—will be explained in greater detail, paying special attention to their unique properties and implementations for several mobile devices. These examples are supposed to gain the readers’ interest for alternative interfaces to large audio collections. Besides, we hope to show that alternative interfaces are feasible for both desktop computers and mobile devices and offer a practical approach to pressing issues in accessing digital collections.