Dali server update

Abstract

The Dali server (http://ekhidna2.biocenter.helsinki.fi/dali) is a network service for comparing protein structures in 3D. In favourable cases, comparing 3D structures may reveal biologically interesting similarities that are not detectable by comparing sequences. The Dali server has been running in various places for over 20 years and is used routinely by crystallographers on newly solved structures. The latest update of the server provides enhanced analytics for the study of sequence and structure conservation. The server performs three types of structure comparisons: (i) Protein Data Bank (PDB) search compares one query structure against those in the PDB and returns a list of similar structures; (ii) pairwise comparison compares one query structure against a list of structures specified by the user; and (iii) all against all structure comparison returns a structural similarity matrix, a dendrogram and a multidimensional scaling projection of a set of structures specified by the user. Structural superimpositions are visualized using the Java-free WebGL viewer PV. The structural alignment view is enhanced by sequence similarity searches against Uniprot. The combined structure-sequence alignment information is compressed to a stack of aligned sequence logos. In the stack, each structure is structurally aligned to the query protein and represented by a sequence logo.

Figure 1.

Outputs from the newly added all-against-all comparison option of the Dali server (manual annotation of internal nodes). Structural dendrogram (A) of 61 selected structures from the PHP and amidohydrolase superfamilies and tryptophan synthase as outgroup. The leaves are linked to structural alignment pages. Branches A–H of the dendrogram are labelled in the correspondence analysis plot (B) of the structural similarity matrix (C). Graphics generated with JSPhyloSVG (19), Microsoft Excel and Plotly (Plotly Technologies Inc. Collaborative data science. Montréal, QC, 2015. https://plot.ly).

Figure 2.

Onthe left, view down the (beta-alpha)-barrel of phosphoesterase 2anuA. Colouring is by structural conservation amongst the members of the PHP superfamily (from blue for the highest through green to red for the lowest conservation). The green barrel strand at the middle left runs in reverse direction compared to the phosphatase and polymerase X members of the PHP superfamily, which have a parallel (beta–alpha) barrel. Onthe right, the same view but this time coloured by sequence conservation. The most highly conserved side chains are shown and reveal the location of the active site. Graphics generated with PV (Protein Viewer).

Figure 3.

Comparison between the sequence motifs of the amidohydrolase and PHP superfamilies. The top two rows represent distant members of the amidohydrolase superfamily (dihydroorotase 3griA and phosphotriesterase 1hzyA (27)). The bottom two rows represent distant members of the PHP superfamily (phosphatase 3dcpA and esterase 2anuA). All segments having two identically conserved amino acids in one or both superfamilies are shown. The structural alignment is anchored on 3griA. There is partial overlap between the sequence motifs but overall the active site is constructed differently. Logos are generated with Skylign from sequence alignments by SANSparallel.