SECONDARY STRUCTURE OF PROTEINS
α-Helix, β-Pleated Sheet, Bends, Ramachandran Plot and Prediction of Secondary Structure
Prepared by
Dr Bhabesh Nath
Assistant Professor
Department of Zoology
B N College Autonomous, Dhubri
1. Introduction
Proteins are linear polymers of amino acids joined by peptide bonds, but their biological activity depends on the three-dimensional shape into which the chain folds. Structure is described at four hierarchical levels. The primary structure is the amino acid sequence. The secondary structure is the regular, local folding of the polypeptide backbone, stabilised mainly by hydrogen bonds between the backbone C=O and N–H groups. The tertiary structure is the overall three-dimensional fold of a single chain, and the quaternary structure is the arrangement of several chains (subunits) in a multimeric protein.
The term secondary structure refers to the local spatial arrangement of main-chain atoms, without regard to the conformation of the side chains or to the relationship with other segments. The common elements are the α-helix, the β-pleated sheet and bends (turns), together with irregular regions called loops or random coil. About 30–35% of residues in a typical globular protein lie in α-helices, about 20–25% in β-strands, and the remainder in turns and loops. The two regular structures were predicted by Linus Pauling, Robert Corey and Herman Branson in 1951, years before the first protein crystal structure (myoglobin, Kendrew, 1958) confirmed them.
1.1 The peptide bond: the basis of secondary structure
The peptide bond (C–N) has about 40% double-bond character because of resonance between the carbonyl oxygen and amide nitrogen. Consequences of this partial double bond are important for understanding every secondary structure:
• The six atoms Cα, C, O, N, H and the next Cα lie in one plane, so the peptide group is rigid and planar.
• The bond length (about 1.32 Å) is shorter than a normal C–N single bond (1.47 Å) and longer than a C=N double bond (1.25 Å).
• Most peptide bonds are in the trans configuration (ω ≈ 180°) because it avoids steric clash between adjacent Cα atoms. The cis form (ω ≈ 0°) is rare except before proline (about 5–6% of X–Pro bonds).
• Free rotation is possible only about the N–Cα bond (phi, φ) and the Cα–C bond (psi, ψ). The whole backbone conformation can therefore be described by the series of φ and ψ angles, which is the idea behind the Ramachandran plot.
2. The α-Helix
The α-helix is the most abundant secondary structure and was proposed by Pauling, Corey and Branson in 1951. The polypeptide backbone is coiled into a rod-like right-handed spiral, with the side chains (R groups) projecting outward and away from the helix axis.
2.1 Structural features
Parameter | Value / description |
Handedness | Right-handed (clockwise) in almost all proteins; left-handed α-helix is sterically unfavourable for L-amino acids |
Residues per turn | 3.6 amino acids |
Pitch (rise per turn) | 5.4 Å |
Rise per residue | 1.5 Å along the helix axis |
Hydrogen bonding | C=O of residue i bonds to N–H of residue i + 4 (parallel to the helix axis) |
H-bonded ring size | 13 atoms, hence also written as the 3.6₁₃ helix |
Dihedral angles | φ ≈ −57°, ψ ≈ −47° |
Helix diameter | About 2.3 Å for the backbone core; about 10 Å including side chains |
2.2 Stabilisation and the helix dipole
• Every peptide bond except those at the ends of the helix takes part in two hydrogen bonds, so the helix is highly stabilised by the cumulative effect of many weak bonds. The first four N–H groups and the last four C=O groups at the ends are free of H-bonds.
• All the peptide dipoles point in the same direction, giving a net helix dipole with a partial positive charge at the N-terminus and partial negative charge at the C-terminus. Negatively charged residues (Asp, Glu) are favoured near the N-terminal end and positively charged residues (Lys, Arg) near the C-terminal end.
• Interactions between side chains three or four residues apart (i, i + 3 and i, i + 4), such as salt bridges and hydrophobic contacts, further stabilise the helix.
2.3 Residues that favour or disfavour the helix
• Strong helix formers: Ala, Glu, Leu, Met, Gln, Lys, Arg.
• Helix breakers: Proline has no amide hydrogen to donate (its N is part of a ring) and its rigid ring restricts φ to about −63°, introducing a kink of about 26° in the axis. Glycine has too much conformational freedom, which destabilises the ordered structure.
• Residues with bulky or branched side chains near the backbone (Val, Ile, Thr) and clusters of like charges (for example many consecutive Glu or Lys) destabilise the helix through steric hindrance and electrostatic repulsion.
2.4 Variants and examples
• 3₁₀ helix: tighter, with 3 residues per turn and H-bonds between i and i + 3; found at the ends of α-helices.
• π helix: wider, 4.4 residues per turn with i to i + 5 H-bonds; rare.
• Amphipathic helix: one face is hydrophobic and the other hydrophilic; common in membrane-associated proteins and in leucine zippers.
• Examples: α-keratin of hair, wool, nails and horn (coiled-coil of two helices, with disulphide cross-links), myosin tail, tropomyosin, and the transmembrane segments of bacteriorhodopsin. Myoglobin is about 75% α-helical.
3. The β-Pleated Sheet
In the β-pleated sheet (Pauling and Corey, 1951) the polypeptide chain is almost fully extended. The segments, called β-strands, are 5–10 residues long. Strands lie side by side and are joined by hydrogen bonds between the C=O and N–H groups of adjacent strands (inter-strand), unlike the intra-chain bonds of the α-helix. The sheet looks pleated, like folded paper, because the Cα atoms sit alternately above and below the plane; the side chains point alternately up and down perpendicular to the sheet.
3.1 Types of β-sheet
Feature | Parallel | Antiparallel | Mixed |
Direction of strands | All strands run N→C in the same direction | Adjacent strands run in opposite directions | Both arrangements in one sheet |
H-bond geometry | Angled (non-linear), longer, weaker | Linear (straight) and short, stronger | Both |
Repeat distance | 6.5 Å | 7.0 Å | – |
φ, ψ (approx.) | −119°, +113° | −139°, +135° | – |
Connection | Needs long crossover connections between strands | Short hairpin (β-turn) loops suffice | – |
Stability | Less stable; usually buried in the core | More stable | – |
3.2 Other features
• Distance between residues along a strand is about 3.3–3.5 Å (compared with 1.5 Å in the helix), showing that the chain is extended.
• Twist: real sheets are not flat; they have a right-handed twist when viewed along the strands, which can coil into β-barrels (for example porins and green fluorescent protein).
• β-bulge: an irregularity caused by an extra residue in one strand that disrupts the H-bond pattern.
• Residues favouring β-strands: Val, Ile, Tyr, Phe, Trp, Thr (large, β-branched or aromatic). Residues disfavouring them: Pro, Asp, Glu, Gly.
• Common motifs: β-hairpin, β-meander, Greek key and β-α-β motif.
• Examples: silk fibroin (antiparallel sheets rich in Gly–Ala/Ser repeats, stacked so the small side chains interdigitate), immunoglobulin domains, concanavalin A, and the fibrils of amyloid in Alzheimer's disease (cross-β structure).
3.3 Comparison of α-helix and β-sheet
Property | α-Helix | β-Pleated sheet |
Backbone | Coiled, rod-like | Extended, zig-zag/pleated |
H-bonds | Intra-chain (i to i + 4), parallel to axis | Inter-strand, perpendicular to strand axis |
Rise per residue | 1.5 Å | 3.3–3.5 Å |
Side chains | Project outward radially | Alternate above and below the sheet |
Typical protein | α-Keratin, myoglobin | Silk fibroin, β-keratin of feathers |
Stretchability | Can be stretched (elastic, converts to β form) | Not stretchable; flexible but inelastic |
4. Bends, Turns and Loops
Globular proteins are compact because the chain reverses direction many times. These reversals are made by short, non-repetitive segments called bends (turns), which often lie on the protein surface and connect helices and strands. About one-third of all residues in globular proteins are in turns and loops.
4.1 β-Turn (reverse turn, β-bend)
• Made of four amino acid residues (i to i + 3). The C=O of residue i forms a hydrogen bond with the N–H of residue i + 3, so the chain turns through 180°. The distance between Cα(i) and Cα(i + 3) is less than 7 Å.
• Type I turn is the most common (about twice as frequent as type II). For residues i + 1 and i + 2 the φ, ψ values are approximately (−60°, −30°) and (−90°, 0°).
• Type II turn has φ, ψ of about (−60°, +120°) and (+80°, 0°). Because the i + 2 position faces steric clash with the side chain in this conformation, it is usually glycine.
• Types I′ and II′ are mirror images, found in tight β-hairpins and favoured by Gly and Asn.
• Proline is common at position i + 1 (its fixed φ suits the turn) and glycine at position i + 2 (its flexibility allows the tight angle). Asp, Asn and Ser are also frequent.
4.2 Other turns and loops
• γ-turn: three residues, H-bond between C=O of residue i and N–H of residue i + 2 (a seven-membered ring); less common than the β-turn.
• α-turn: five residues; π-turn: six residues.
• Ω (omega) loops: longer loops of 6–16 residues in which the ends are close together, giving a Greek-letter Ω shape; often involved in ligand binding and recognition.
• β-hairpin: two antiparallel strands joined by a short turn; the commonest connection in antiparallel sheets.
Turns have functional significance: they allow the chain to fold compactly, are often sites of antigenic determinants and enzyme-active-site loops, and are exposed to solvent because they are rich in polar and charged residues.
5. The Ramachandran Plot
The Ramachandran plot was introduced in 1963 by G. N. Ramachandran, C. Ramakrishnan and V. Sasisekharan at the University of Madras. It is a two-dimensional map of the backbone dihedral angles φ (phi) on the x-axis and ψ (psi) on the y-axis, each from −180° to +180°. It shows which combinations of φ and ψ are sterically allowed for an amino acid residue in a polypeptide.
5.1 Definition of the angles
• φ (phi): rotation about the N–Cα bond, defined by the atoms C(i−1)–N(i)–Cα(i)–C(i).
• ψ (psi): rotation about the Cα–C bond, defined by N(i)–Cα(i)–C(i)–N(i+1).
• ω (omega): rotation about the peptide C–N bond, fixed at about 180° (trans) because of the partial double-bond character.
• By convention, when the backbone is fully extended with all atoms in one plane, φ = ψ = 180° (or ±180°).
5.2 Principle
Using a hard-sphere model with van der Waals radii, Ramachandran calculated which φ–ψ combinations cause atoms to approach closer than allowed (steric clash). Most combinations are forbidden. For a typical L-amino acid only about 10–20% of the plot is accessible, which is why a polypeptide cannot adopt a random conformation freely and why folding is constrained.
Figure 1. Schematic Ramachandran plot showing the major allowed regions (approximate, not to scale).
5.3 Regions of the plot
Region | Approx. (φ, ψ) | Structure and remarks |
Upper-left quadrant (core β region) | (−120°, +130°) | β-strands and β-sheets (parallel and antiparallel); extended conformation |
Lower-left quadrant | (−57°, −47°) | Right-handed α-helix; the largest and most populated allowed region |
Polyproline II region | (−75°, +145°) | Extended left-handed helix of collagen and polyproline; also common in unfolded segments |
Upper/lower right quadrant (small island) | (+57°, +47°) | Left-handed α-helix (αL); rare, mainly Gly, Asn and Asp |
Remaining areas | – | Disallowed due to steric hindrance for non-glycine residues |
5.4 Special cases
• Glycine has only a hydrogen as its side chain, so it has much less steric hindrance and occupies a much larger area of the plot, including regions forbidden to other residues (both the left and right sides, roughly symmetrical about the origin).
• Proline has a ring linking N and Cα, which fixes φ at about −63° ± 15°, so it occupies a narrow band of the plot.
• Residues with β-branched side chains (Val, Ile, Thr) are more restricted and favour the β-region.
• Pre-proline residues also have a narrower allowed region than other non-glycine residues.
5.5 Applications
• Validation of protein structures determined by X-ray crystallography, NMR or modelling. In a good-quality structure more than 90% of non-glycine, non-proline residues lie in the most favoured regions (as assessed by programs such as PROCHECK and MolProbity). Outliers point to errors or to strained, functionally important residues.
• Identifying secondary structure of a protein: clusters in the α and β regions indicate helical and sheet content.
• Protein design and modelling: limits the conformational space explored in folding simulations and loop modelling.
• Comparing conformations of mutants and homologues, and understanding the structural effects of Gly/Pro substitutions.
6. Prediction of Secondary Structure
Determining a structure experimentally by X-ray crystallography, NMR or cryo-EM is slow and expensive, whereas millions of sequences are known. Secondary structure prediction aims to predict, from the amino acid sequence alone, whether each residue is in a helix (H), strand (E) or coil/turn (C). It is an important intermediate step in tertiary structure prediction, fold recognition, domain identification and protein engineering.
6.1 Basis of prediction
• Different amino acids have different intrinsic propensities for helix, strand and turn.
• Secondary structure is determined largely by local sequence, but also depends on longer-range interactions; this limits accuracy.
• Homologous proteins have conserved structure, so evolutionary information (multiple sequence alignments and profiles) improves predictions greatly.
6.2 Generations of methods
First generation – single-residue statistics (1970s). These use the propensity of each amino acid for each structure, giving about 50–60% accuracy.
• Chou–Fasman method (1974): based on empirical propensity parameters Pα, Pβ and Pt (turn) calculated from the frequency of each amino acid in known structures. Examples: Glu (Pα ≈ 1.51), Met (1.45), Ala (1.42) and Leu (1.21) are strong helix formers; Val (Pβ ≈ 1.70), Ile (1.60), Tyr (1.47) and Phe (1.38) are strong strand formers; Pro and Gly are strong turn formers.
Algorithm (outline): (1) assign the three propensities to every residue; (2) find helix nuclei, a cluster of 6 residues of which at least 4 have Pα > 1.03; extend the region in both directions until a set of four consecutive residues has an average Pα below 1.03; (3) find strand nuclei, a cluster of 5 residues of which at least 3 have Pβ > 1.05, and extend similarly; (4) where regions overlap, assign the one with the higher average propensity; (5) locate turns using a four-residue window with high turn propensity (Pt) and the position-specific frequencies f(i), f(i+1), f(i+2), f(i+3).
Second generation – segment statistics and information theory (late 1970s–1980s). These consider a window of neighbouring residues, reaching about 60–65% accuracy.
• GOR method (Garnier, Osguthorpe and Robson, 1978): uses information theory and Bayesian statistics. The conformation of a residue is predicted from the information contributed by itself and by each of the 8 residues on either side (a window of 17 residues). Later versions (GOR IV) use pair frequencies and reach about 64–65% accuracy.
• Lim's method uses stereochemical rules; the Robson/Garnier approaches belong to this class too.
Third generation – evolutionary profiles and machine learning (1990s onward). These use multiple sequence alignments and artificial neural networks, reaching about 70–80% accuracy.
• PHD (Rost and Sander, 1993): neural networks fed with multiple-alignment profiles; first method to exceed 70%.
• PSIPRED (Jones, 1999): two-stage neural network applied to the position-specific scoring matrix from PSI-BLAST; about 80% accuracy (Q3).
• JPred, SOPMA, PROFsec, SPIDER and DeepCNF/deep-learning predictors are other widely used tools; consensus servers combine multiple methods.
Recent advances. Deep learning methods, culminating in AlphaFold2 (Jumper et al., 2021), predict full 3-D structure with near-experimental accuracy for many proteins, from which secondary structure is read directly (using programs such as DSSP). Dedicated secondary structure predictors now approach 84–86% accuracy, close to the practical limit of about 88–90% set by differences in how experts assign structures.
6.3 Assessment of accuracy
• Q3 score: percentage of residues correctly predicted in the three states (H, E, C).
• SOV (segment overlap): measures how well entire segments are predicted, not only individual residues.
• Reference assignments of true structure from known 3-D coordinates are made with DSSP (Kabsch and Sander, 1983, based on H-bond energies) or STRIDE.
6.4 Comparison of methods
Method | Principle | Input | Approx. accuracy |
Chou–Fasman | Residue propensities, nucleation and extension rules | Single sequence | 50–60% |
GOR | Information theory, 17-residue window | Single sequence | 60–65% |
PHD | Neural network with alignment profile | Multiple alignment | 70–72% |
PSIPRED / JPred | Neural networks on PSI-BLAST profiles | Profile from database search | ~80% |
Deep learning, AlphaFold2 | Deep neural networks with co-evolution | Sequence and alignments | 84–86% or more (structure-based) |
6.5 Limitations
• Accuracy is lowest for β-strands and for the ends of segments (exact boundaries) because β-strands depend on long-range contacts.
• Same sequence can adopt different structures depending on context (chameleon sequences), and intrinsically disordered regions are difficult to predict.
• Orphan proteins with few homologues have poor profiles, which reduces accuracy.
7. Quick Summary
• Secondary structure = local, regular folding of the backbone held by H-bonds between C=O and N–H groups; the main types are α-helix, β-sheet and turns.
• α-helix: right-handed, 3.6 residues/turn, 5.4 Å pitch, H-bond i → i + 4, φ ≈ −57°, ψ ≈ −47°; broken by Pro.
• β-sheet: extended strands, inter-strand H-bonds, parallel or antiparallel, 3.3–3.5 Å per residue.
• β-turn: four residues, H-bond i → i + 3, rich in Pro and Gly.
• Ramachandran plot: φ vs ψ map; allowed regions from steric limits; used to validate structures; Gly broad, Pro narrow.
• Prediction: Chou–Fasman (50–60%) → GOR (≈65%) → neural networks with profiles (70–80%) → deep learning (>84%).
8. Multiple Choice Questions (10)
Q1. Who proposed the α-helix and β-pleated sheet structures in 1951?
(a) Watson and Crick
(b) Pauling, Corey and Branson
(c) Sanger and Edman
(d) Kendrew and Perutz
Q2. The number of amino acid residues per turn in a typical α-helix is:
(a) 2.0
(b) 3.0
(c) 3.6
(d) 4.4
Q3. In an α-helix the C=O group of residue i forms a hydrogen bond with the N–H group of residue:
(a) i + 2
(b) i + 3
(c) i + 4
(d) i + 7
Q4. Which amino acid is known as a 'helix breaker' because it lacks an amide hydrogen and has a rigid ring?
(a) Alanine
(b) Leucine
(c) Proline
(d) Glutamate
Q5. In antiparallel β-sheets, compared with parallel sheets, the hydrogen bonds are:
(a) Longer and angled
(b) Shorter and nearly linear
(c) Absent
(d) Only intra-strand
Q6. A β-turn is made up of how many amino acid residues?
(a) Two
(b) Three
(c) Four
(d) Six
Q7. In the Ramachandran plot, φ is the rotation about the:
(a) Cα–C bond
(b) N–Cα bond
(c) C–N peptide bond
(d) Cα–Cβ bond
Q8. Which amino acid has the widest allowed region in the Ramachandran plot?
(a) Valine
(b) Proline
(c) Glycine
(d) Isoleucine
Q9. The Chou–Fasman method of secondary structure prediction is based on:
(a) Neural networks
(b) Residue propensities for each structure
(c) Multiple sequence alignment profiles
(d) Molecular dynamics
Q10. Which program/server uses PSI-BLAST profiles with neural networks to predict secondary structure?
(a) GOR I
(b) Chou–Fasman
(c) PSIPRED
(d) DSSP
Answer Key
Q1 | Q2 | Q3 | Q4 | Q5 | Q6 |
(b) | (c) | (c) | (c) | (b) | (c) |
Q7 | Q8 | Q9 | Q10 |
(b) | (c) | (b) | (c) |
9. Exam-Oriented Questions
A. One-Mark Questions (with answers)
1. Define secondary structure of a protein.
Ans: The regular, local folding of the polypeptide backbone (such as α-helix and β-sheet) stabilised by hydrogen bonds between backbone C=O and N–H groups.
2. What is the pitch of the α-helix?
Ans: 5.4 Å (3.6 residues × 1.5 Å).
3. Name the bond that gives rigidity and planarity to the peptide unit.
Ans: The peptide (C–N) bond, which has partial double-bond character.
4. Which amino acid disrupts an α-helix?
Ans: Proline (also glycine).
5. What type of hydrogen bonds stabilise a β-sheet?
Ans: Inter-strand hydrogen bonds between C=O and N–H groups of adjacent strands.
6. Give one example of a protein rich in β-sheets.
Ans: Silk fibroin.
7. Which residue is commonly found at position i + 2 of a type II β-turn?
Ans: Glycine.
8. What do the axes of the Ramachandran plot represent?
Ans: The backbone dihedral angles φ (x-axis) and ψ (y-axis).
9. Expand DSSP and state its use.
Ans: Define Secondary Structure of Proteins; assigns secondary structure from 3-D coordinates based on hydrogen-bond patterns.
10. Name two methods used to predict secondary structure.
Ans: Chou–Fasman and GOR (also PSIPRED, PHD).
B. Two-Mark Questions (with answers)
1. Differentiate between parallel and antiparallel β-sheets.
Ans: In parallel sheets all strands run in the same N→C direction and the H-bonds are angled and weaker (repeat 6.5 Å). In antiparallel sheets adjacent strands run in opposite directions and the H-bonds are straight, shorter and stronger (repeat 7.0 Å).
2. Why is proline called a helix breaker?
Ans: Proline's nitrogen is part of a ring, so it has no amide hydrogen to form the i → i + 4 H-bond, and the ring restricts φ to about −63°, which introduces a kink in the helix.
3. What is a β-turn? Mention its significance.
Ans: A four-residue bend (i to i + 3) in which the C=O of residue i bonds with the N–H of residue i + 3, reversing the chain direction by 180°. It allows compact folding and connects strands in β-hairpins; usually Pro at i + 1 and Gly at i + 2.
4. State two applications of the Ramachandran plot.
Ans: (1) Validating the stereochemical quality of experimentally determined or modelled protein structures; (2) identifying secondary structure content and unusual (outlier) residues.
5. Why does glycine occupy a larger area in the Ramachandran plot?
Ans: Its side chain is only a hydrogen atom, so steric hindrance is minimal and a wider range of φ and ψ values is allowed.
6. What is the Q3 score?
Ans: The percentage of residues whose secondary structure state (helix, strand or coil) is correctly predicted by a prediction method.
7. List the dimensions of the α-helix.
Ans: 3.6 residues per turn, pitch 5.4 Å, rise 1.5 Å per residue, φ ≈ −57°, ψ ≈ −47°.
8. What is a helix dipole?
Ans: The cumulative dipole of aligned peptide bonds in an α-helix, giving a partial positive charge at the N-terminus and a partial negative charge at the C-terminus.
C. Five-Mark Questions (answer outlines)
1. Describe the structure of the α-helix and the forces stabilising it.
Key points for answer:
– Proposed by Pauling, Corey and Branson (1951); right-handed spiral with side chains outward.
– 3.6 residues per turn, pitch 5.4 Å, rise 1.5 Å per residue; φ ≈ −57°, ψ ≈ −47°.
– H-bond between C=O of residue i and N–H of residue i + 4, parallel to the axis; 13-atom ring.
– Stabilised by H-bonds, van der Waals packing of the backbone and i, i + 3/i + 4 side-chain interactions; helix dipole.
– Helix formers (Ala, Leu, Glu, Met) and breakers (Pro, Gly); examples: α-keratin, myoglobin.
2. Explain the β-pleated sheet and compare its parallel and antiparallel forms.
Key points for answer:
– Extended strands (5–10 residues) arranged side by side, with pleated appearance.
– Inter-strand H-bonds between C=O and N–H groups; side chains alternate above and below the sheet.
– Parallel: same direction, angled bonds, repeat 6.5 Å; antiparallel: opposite direction, linear bonds, repeat 7.0 Å; mixed also exist.
– Rise of 3.3–3.5 Å per residue; right-handed twist; formers Val, Ile, Tyr, Phe, Trp.
– Examples: silk fibroin, immunoglobulin domains, β-barrels; draw a labelled diagram.
3. Write a note on bends (β-turns) and their types.
Key points for answer:
– Chain-reversing segments on the surface joining helices and strands; four residues in β-turns, Cα(i)–Cα(i + 3) < 7 Å.
– H-bond between C=O(i) and N–H(i + 3).
– Type I (commonest) and Type II (needs Gly at i + 2); mirror types I′ and II′.
– Pro at i + 1, Gly at i + 2; also Asn, Asp, Ser.
– Other turns: γ-turn (3 residues), α-turn, Ω loops and β-hairpins; functional importance.
4. Explain the Ramachandran plot and its significance.
Key points for answer:
– Introduced by Ramachandran, Ramakrishnan and Sasisekharan (1963); plot of φ against ψ.
– Definition of φ (N–Cα) and ψ (Cα–C); ω fixed at 180°.
– Allowed regions from van der Waals radii: β-sheet (upper left), right-handed α-helix (lower left), left-handed helix (small, upper right), PPII.
– Special cases: Gly broad, Pro narrow (φ ≈ −63°).
– Uses: structure validation (>90% in favoured regions), identifying outliers, modelling; include sketch.
5. Discuss the methods for prediction of protein secondary structure.
Key points for answer:
– Principle: amino acid propensities and evolutionary information.
– Chou–Fasman: Pα, Pβ, Pt; helix nucleation (4 of 6 residues with Pα > 1.03), strand nucleation (3 of 5 with Pβ > 1.05); accuracy 50–60%.
– GOR: information theory, 17-residue window; about 65%.
– Neural networks: PHD, PSIPRED, JPred with profiles; 70–80%.
– Deep learning/AlphaFold2; assessment by Q3 and SOV; limitations.
D. Additional Practice Questions
1. Distinguish between the α-helix and β-sheet in a tabular form.
2. Why do globular proteins contain β-turns? Describe the geometry of a type I β-turn.
3. Explain how a Ramachandran plot is used to assess a modelled protein structure.
4. Compare first-, second- and third-generation methods of secondary structure prediction.
5. What is the effect of Gly and Pro residues on secondary structure? Explain with reference to the Ramachandran plot.
10. References
1. Nelson DL, Cox MM. Lehninger Principles of Biochemistry, 8th ed. W.H. Freeman; 2021.
2. Voet D, Voet JG. Biochemistry, 4th ed. John Wiley & Sons; 2011.
3. Berg JM, Tymoczko JL, Gatto GJ, Stryer L. Biochemistry, 9th ed. W.H. Freeman; 2019.
4. Branden C, Tooze J. Introduction to Protein Structure, 2nd ed. Garland Science; 1999.
5. Lesk AM. Introduction to Protein Science: Architecture, Function and Genomics, 3rd ed. Oxford University Press; 2016.
6. Pauling L, Corey RB, Branson HR. The structure of proteins: two hydrogen-bonded helical configurations of the polypeptide chain. Proc Natl Acad Sci USA. 1951;37(4):205–211.
7. Pauling L, Corey RB. The pleated sheet, a new layer configuration of polypeptide chains. Proc Natl Acad Sci USA. 1951;37(5):251–256.
8. Ramachandran GN, Ramakrishnan C, Sasisekharan V. Stereochemistry of polypeptide chain configurations. J Mol Biol. 1963;7:95–99.
9. Venkatachalam CM. Stereochemical criteria for polypeptides and proteins. V. Conformation of a system of three linked peptide units. Biopolymers. 1968;6:1425–1436.
10. Chou PY, Fasman GD. Prediction of protein conformation. Biochemistry. 1974;13(2):222–245.
11. Garnier J, Osguthorpe DJ, Robson B. Analysis of the accuracy and implications of simple methods for predicting the secondary structure of globular proteins. J Mol Biol. 1978;120:97–120.
12. Kabsch W, Sander C. Dictionary of protein secondary structure: pattern recognition of hydrogen-bonded and geometrical features. Biopolymers. 1983;22:2577–2637.
13. Rost B, Sander C. Prediction of protein secondary structure at better than 70% accuracy. J Mol Biol. 1993;232:584–599.
14. Jones DT. Protein secondary structure prediction based on position-specific scoring matrices. J Mol Biol. 1999;292:195–202.
15. Jumper J, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596:583–589.
16. Lovell SC, et al. Structure validation by Cα geometry: φ, ψ and Cβ deviation. Proteins. 2003;50:437–450.
0 Comments