The bioinformatic pipeline that predicts the 3D native-like conformation of proteins using its amino acid sequence, is known as Protein Structure Prediction. Since, the structural features of proteins can shed light on their biological functions, it is important to analyze the protein structures.
Methods of Protein structure prediction
Experimental methods
Chiefly, there are three experimental methods for protein structure prediction, these are as follows:
X-ray Crystallography - it is a wet lab technique that is being utilized in proteomics for the determination of protein structures in their crystalline form. The basis of this technique lies on the fact that the crystalline atoms cause the x-ray beam to diffract at different angles and in different directions. Hence, by measuring the angles at which the beam diffracts and the intensities of the diffracting beam by a detector (known as Crystallographer), the 3D image of the densities of the crystal’s elements can be predicted. In this way the bond angles, bond lengths and other information as well to predict the 3D structure of the protein molecule.
NMR - Nuclear Magnetic Resonance is a wet lab technique used for the determination of molecular structures of biological samples and for the analytical studies and purity identification of different samples in chemistry. In this technique, samples are placed in a strong magnetic field. When these samples are exposed to such a high level magnetic field, the nuclei within the atoms of the sample start resonating which causes the transition effect due to discrete energy levels. Such energy gaps are then measured and visualized in the form of spectras. The data gathered in this process is then utilized to explain the structure of the sample chemically.
Electron-Cryo Microscopy - it is a versatile tool in the analysis of structures of proteins and biological macromolecular assemblies. Such latest advances in the field of protein structure analysis provide exciting opportunities for the three-dimensional structural determination of macromolecular complexes that are either too large or too heterogeneous to be investigated by conventional X-ray crystallography or NMR. Understanding the function of protein or macromolecular complexes is often helped by combining data from electron microscopy and X-ray crystallography.
Computational methods
These can be of two types:
1. Knowledge based - the protein structure prediction methods based on the knowledge of already existing proteins information present in databases such as PDB, UniProt, etc. These include the following two approaches:
a. Homology Modeling - also known as Comparative modeling - is a technique referred to the construction of a 3D structure model of target protein from its amino acid sequence and an experimental 3D structure of a related homologous protein (template sequence). It is also referred to as “Template-based Modeling”.
Homology modeling is based on the knowledge that evolutionary related proteins share similar structure. This approach utilizes the fact that structural conformation of proteins is more conserved than their sequence.
Tools for homology modeling include IntFOLD, RaptorX, HHpred, MODELLER, MOE, Phyre/Phyre2, ROBETTA, Swiss-Model, Prime, and FoldX, etc.
b. Threading - also known as fold recognition method - predicts the structural folds of an unknown protein sequence by connecting the sequence into a structural database and selecting the best-fit model. This comparison highlights the most suitable matchings of secondary structures, which are highly conserved evolutionarily.
Tools for protein threading include I-TASSER, RaptorX, IntFOLD, FALCON, MUSTER, Phyre/Phyre2, DeepFR, SPARKS-X, BBSP, and GenTHREADER, etc.
2. Ab initio - as the name suggests, this approach predicts the protein structure from the very scratch (de novo). When there is no suitable template structure available ab initio methods can be utilized for the protein structure prediction using the linear amino acid sequence of the target protein. The structure is predicted based on the global free energies of the protein models.
Tools for ab initio modeling of target proteins include I-TASSER, Rosetta, ROBETTA, FALCON, Abalone, PEP_FOLD, trRosetta, etc.
Bioinfolytics & Our Services
If you want to learn and develop your expertise in Protein Structure Prediction & analysis or if you’ve any research project that requires such services, join our Gray Bioinformatics plans from BioCode, where we’re providing you with complete video lectures on the databases and the tools that are required for this purpose and teach you how to analyze the results and draw logical conclusions and hypothesis from the results.
To join our Gray Bioinformatics plans, visit us at https://www.biocode.ltd/ and enroll yourself to develop your skills in Bioinformatics databases and tools at affordable costs.
Protein structure prediction is a series of techniques in bioinformatics to predict the structure of a protein sequence from only its amino acid sequence. Through BioinfoLytics, our team is very happy to help you to provide the services for 3D structure prediction, secondary structure prediction, and more.
Through BioinfoLytics, we’re providing our expertise in these protein structure prediction techniques (i.e., Ab Initio, Homology Modeling, Threading) and can predict protein structure and then evaluate those structures as well from your provided amino acid sequences.
We’ll be delighted to provide you our platform of BioinfoLytics, where you can sell your skills as a freelancer.
For further information on our services, visit us at https://www.biocode.ltd/bioinfolytics
Or directly contact us at bioinfolytics@biocode.ltd