Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 2, e456

A Complete, Multi-Level Conformational Clustering of Antibody Complementarity-Determining Regions


A Complete, Multi-Level Conformational Clustering of Antibody Complementarity-Determining Regions

Dimitris Nikoloudis et al. PeerJ.


Classification of antibody complementarity-determining region (CDR) conformations is an important step that drives antibody modelling and engineering, prediction from sequence, directed mutagenesis and induced-fit studies, and allows inferences on sequence-to-structure relations. Most of the previous work performed conformational clustering on a reduced set of structures or after application of various structure pre-filtering criteria. In this study, it was judged that a clustering of every available CDR conformation would produce a complete and redundant repertoire, increase the number of sequence examples and allow better decisions on structure validity in the future. In order to cope with the potential increase in data noise, a first-level statistical clustering was performed using structure superposition Root-Mean-Square Deviation (RMSD) as a distance-criterion, coupled with second- and third-level clustering that employed Ramachandran regions for a deeper qualitative classification. The classification of a total of 12,712 CDR conformations is thus presented, along with rich annotation and cluster descriptions, and the results are compared to previous major studies. The present repertoire has procured an improved image of our current CDR Knowledge-Base, with a novel nesting of conformational sensitivity and specificity that can serve as a systematic framework for improved prediction from sequence as well as a number of future studies that would aid in knowledge-based antibody engineering such as humanisation.

Keywords: Antibody structure; CDR conformation; Canonical model; Clustering; Dynamic hybrid tree-cut; Humanisation; Nested architecture; Prediction; Redundant repertoire.


Figure 1
Figure 1. Superposition of 7-residue and 11-residue CDR-L2.
The 5 C-terminal residues of 1A4K (in red) 7 residue CDR-L2 (L52–L56) are superposed to the equivalent portion of 3FFD (in blue) 11 residue CDR-L2. Position L51 is highlighted in green, as the best insertion point in the structural numbering scheme. Graphics created with Swiss-PdbViewer (
Figure 2
Figure 2. Superposition of Light Framework 3 with an insertion onto a typical LFR3.
Residues L60–L75 of crystal structure 1PW3 (in red), containing an insertion, are superposed onto a typical example of the equivalent Light chain fragment (here 1A4K, in blue). The new insertion point was introduced in position L67 (highlighted in green). Graphics created with Swiss-PdbViewer (
Figure 3
Figure 3. Superposition of Heavy Framework 3 with an insertion onto a typical HFR3.
The Cα-trace of a two-leg superposition of residues H65–H73 and H76–H78 of crystal structures 3RPI (in yellow) and 3SE8 (in red), containing an insertion, onto the equivalent residues of a typical structure without an insertion (here 3MLY, in blue). The proposed insertion point H74 is highlighted in green in 3MLY and is shown with its side chain (Ser). Graphics created with Swiss-PdbViewer (
Figure 4
Figure 4. Ramachandran plot divided into conformational regions.
A: α-helix region; B: β-sheet region; D: δ-region; G: γ-region; L: left-handed helix region; P: polyproline II region. For the construction of reduced-Ramachandran logos (r-RL), residues belonging to regions with similar conformations were represented by the same letter: (A/D) = A, (B/P) = B, (L/G) = L. For the construction of full-Ramachandran logos (f-RL), each conformational region was represented individually. E.g., Ramachandran logos for CDR-L3 1TJH_L:r-RL: BBAABBBBB f-RL: BBDABPPPB.
Figure 5
Figure 5. Example of the nested clusters architecture.
Level-1 cluster H1-13-III (i.e., the third top-level cluster of 13-residues CDR-H1), defined by RMSD-based hierarchical clustering, contains 3 Level-2 clusters, the members of each sharing the same reduced-Ramachandran logo, and in total 11 Level-3 clusters, the members of each sharing the same full-Ramachandran logo. All Level-3 clusters share the same reduced-Ramachandran logo with their parent Level-2 cluster, but each one displays a distinct full-Ramachandran logo.
Figure 6
Figure 6
Illustration of the parameters taken into account for the dendrogram pruning of CDR-L1/12 residues with the Dynamic Hybrid method. The minimum gap statistic (gmin) defines the minimum required distance between the average core scatter and the joining height of the clusters (‘Gap’), for successful cluster formation. In this example, gmin is set lower than the displayed Gaps, so nodes above its value were considered as different clusters.
Figure 7
Figure 7
A comparative view of all CDR-H1/13 residue clusters obtained in this work (in yellow), superposed to their correspondences from North, Lehmann & Dunbrack (2011), where applicable. Level-1 clusters from this work expand whenever possible towards closely-related variants, which are then further classified at levels 2 and 3 (complete 3-level classification in this work of the external median is given in brackets). This can be appreciated in clusters H1-13-I and H1-13-III from this work. The last four structures of this figure correspond to cluster medians from North, Lehmann & Dunbrack (2011) that were classified as outliers/singletons in this work.

Similar articles

See all similar articles

Cited by 3 PubMed Central articles


    1. Al-Lazikani B, Lesk AM, Chothia C. Standard conformations for the canonical structures of immunoglobulins. Journal of Molecular Biology. 1997;273:927–948. doi: 10.1006/jmbi.1997.1354. - DOI - PubMed
    1. Barré S, Greenberg AS, Flajnik MF, Chothia C. Structural conservation of hypervariable regions in immunoglobulins evolution. Nature Structural Biology. 1994;1:915–920. doi: 10.1038/nsb1294-915. - DOI - PubMed
    1. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN, Bourne PE. The protein data bank. Nucleic Acids Research. 2000;28:235–242. doi: 10.1093/nar/28.1.235. - DOI - PMC - PubMed
    1. Chothia C, Lesk AM. Canonical structures for the hypervariable regions of immunoglobulins. Journal of Molecular Biology. 1987;196:901–917. doi: 10.1016/0022-2836(87)90412-8. - DOI - PubMed
    1. Chothia C, Lesk AM, Levitt M, Amit AG, Mariuzza RA, Phillips SEV, Poljak RJ. The predicted structure of immunoglobulin D1.3 and its comparison with the crystal structure. Science. 1986;233:755–758. doi: 10.1126/science.3090684. - DOI - PubMed

Grant support

The authors declare no external funding sources.

LinkOut - more resources