secondary structure


The secondary structure is biochemical and structural biology refers to a biogenic molecule such as protein and nucleic acid (DNA or RNA), a three-dimensional form of partial segment. . It does not describe any particular atomic location, which will be processed in the third level structure.


Due to the hydrogen bond associated with other structural features, the secondary structure defined is slightly unfair. As the protein spiral, the main chain is used in the specific area of ​​the Raman strong syndrome. In this way, it is called "spiral" in this two-faced range, regardless of whether it is a true hydrogen bond. Other slightly unformal definitions are also recommended, and are mostly the concept of application curve differentiation, such as curvature and temporality. The most inconspicuous, it is necessary to calculate the structural biological behavior to determine and record the secondary structure of the atomic level.

The secondary structure of the biological giant molecule can be preliminary to estimate. For proteins, it is known as a common method (wavelength 170-250 nm) round chromatography. The α-helical structure can be displayed in the minimum of 208 nm and 222 nm, while the minimum 204 nm or 207 nm can be displayed in an orthodox or β folding structure, respectively. A less use method is an infrared spectrum that detects the oscillation of the amine group due to hydrogen bonding. Finally, the secondary structure can be accurately moved by the chemical disgins of the nuclear magnetic resonance.


DSSP is an abbreviation of "DefineSecondaryStructureOfproteins", which is a two-dimensional secondary structure that is known to have known three-dimensional structures. The DSSP number is typically description using a single English letter to describe the secondary structure of the protein. The secondary structure is specified based on hydrogen bond mode.

* g: 3 corner spiral (i.e., 310 spiral). The shortest length is 3 residues.

* h: 4 corner spiral (α helix). The shortest length is 4 residues.

* i: 5 corner spiral (π helix). The shortest length is 5 residues.

* t: hydrogen key angle (3, 4 or 5 corners).

* E: Parallel β fold, or / and anti-parallel folding form (extension chain). The shortest length is 2 residues.

* b: residue within a separate β bridge (a pair of beta folding hydrogen bonds)

* S: bending (specified by the unique non-hydrogen bond)

All residues of the above forms are specified in DSSP, and sometimes C, which represents C. represents the ring. Spirals (ie g, h and i) and folding forms require a certain length. This means that the residues adjacent to the primary structure must form the same hydrogen bond mode. If the hydrogen bond mode of the spiral or folding is too short, it will be encoded in t or b, respectively. There are other protein secondary structural numbers, but less use.

Protein secondary structure Prediction

Early protein secondary structure prediction method is to build a tendency based on amino acid forming spiral or folding, and sometimes it must be estimated to form a secondary structure. The method is used. These methods can have an accuracy of about 60% of the three states (helix, folding or curl) of the residue, and the accuracy can be greatly increased to 80% if the multi-sequence alignment can be used. Multi-sequence alignment can know the intensive distribution of amino acids in a certain location (including in its vicinity, 7 residues in each side), while evolutionary processes provide a structural trend more clear drawing. For example, the glymine acid in a certain location in the protein itself has shown that it is an arbitrary shape. However, multi-sequence comparisons can be found that in 95% of the protein after approaching 5 million years, it is a favorable spiral amino acid. Furthermore, the average hydrophobicity is detected in that position, and it will also be found that its residual solubility is consistent with the alpha helix. In combination, these factors show that the original protein-in-glymonic acid is a helix structure, not any arbitrary. A variety of methods combine existing data to form three state predictions, these methods have neural networks, hidden Markov models and support vector machines. The modern prediction method can also provide a trolley score at the prediction result of each position.

The secondary structure prediction method has been constantly calibrated, such as EVA experiments. Based on the test of about 270 weeks, the most accurate approach is Psipred, SAM, Porter, PROF, and SABLE. Interestingly, finding consensus or consistency in these methods, do not enhance their accuracy. The largest improvement is seemingly predicted in beta stocks because the methods used will ignore some β-shares. In general, the highest predictive accuracy can only reach 90%, due to the nature of the DSSP standard method, against the prediction of calibration.

Accurate secondary structural prediction is an important racemic of the three-level structure prediction. For example, a determined βαββ-secondary structural mode is a marker of iron oxidative protein.


The secondary structure of the protein comprises the interaction between the local residue is adjusted by the hydrogen bond. The most common secondary structure is alpha-spiral and β-fold, in addition to β-angle and random curvature. Other spirals, such as 310 spirals, such as 310 spiral, such as 310 spirals, have an advantageous hydrogen bond mode, but these spirals are very rare in natural proteins, and α is helically adversely adversely adversely adversely filled It can only be found in the end. Tighten the corner, loose and flexible circular links more "rules" secondary structures. The arbitrary shape is not a real secondary structure, but it is a class of secondary structures lacking rules.

The amino acid has different capabilities in forming different secondary structures. Proline and glycine will appear on the corner, and can dissert the rule form of the α spiral skeleton, but both have abnormal morphological capabilities. The amino acid having a spiral form is used in the protein, alanine, leucine, glutamate and lysine (amino acid monohydrate "); opposite, large aromatic residues (tryptophan, The amino acids (isobiline, proline and threonine) of tyrosine and phenylalanine) and Cβ are used in the form of a β fold. However, if you are not sufficient to form a reliable method to predict the secondary structure.

Nucleic acid

Nucleic acids also have secondary structures, and most of them are single-stranded ribonucleic acid (RNA) molecules. The RNA secondary structure can be divided into a helical (completed base pair) and different types of rings (unsuccessful nucleotides surrounded by the spiral). The stem ring structure is a base pair of helical structures, and the end is a short-term ring. This stem ring structure is very common and is a basic unit that constructs large structural primitives such as a clover structure (i.e., four screw joints in transport RNA). The inner ring structure (short-free bases in the long base pair of spirals) and bulging (additional insertion in the helix stocks, but there is no pairing base in the relatively shares). Finally, puppet and basetriples will also appear in RNA.

Due to the almost all of the RNA secondary structure, it is made of base pair as an intermediary, which can be said to be determined which base is paired in a molecule or complex. However, traditional Huatan-Krick baseline is not only a matching method of RNA, and the Hob's pairing method is also very common.

Level Structure

Biological information is used in which one application is to use predicted RNA secondary structures to search for genome used as RNA functional forms rather than encoded. For example, the small molecule RNA has a long stem ring structure that is interrupted by a small inner ring. Calculating the possible RNA secondary structure can be used by a dynamic planning method, but it cannot detect a pseudo-knot or other base pairs with a general-purpose method without comprehensive online. There is a random context. Mfold is a website that uses dynamic planning.


Protein and RNA secondary structure can be used in assisting multi-sequence comparison. This alignment can become more accurate after the second-level structural data is added. However, sometimes it is not useful to RNA, which is because the RNA base contrast sequence is more highly stored. Some proteins that cannot be compared to the primary structure, the secondary structure can sometimes find the relationship between them.

Related Articles