HowTo

Single Analysis

Submit a structure file in the pdb format and our server will calculate the lowest frequency normal modes of your molecule. You will then be offered different types of calculations to analyse the modes that were calculated.

In the field for chains, you can specify a chain found in the structure, e.g. "A" or multiple chains in the format "A,B,C". This allows you to analyse specific parts of the protein that you wish to analyse without modifying the structure file. This can be left blank if the entire structure is being analysed.

The checked Perform default analyses box will run the Atomic displacement and Mode Visualisation analyses along with the initial calculations.

For the overlap analysis, you can compare the calculated normal modes of your submitted structure with a different structure of the same protein in a different conformation. The structure provided for the overlap analysis must have the same number of CA-atoms as the one the normal modes were calculated for, and the atoms of the two structures should correspond to each other in the order they appear in the PDB-files. The submission procedure is similar to the above, where you may either choose to use a PDB ID or upload a coordinate file, and specify chain(s) you wish to analyse.

For larger submissions, you can provide your e-mail address so that a link to the results can be sent to you when they are ready. The size limit for a single submission is 20000 amino acids.

Comparative Analysis

You need to prepare a file containing your sequence alignment of the proteins you wish to compare and a protein structure files for each of the proteins.
The alignment file needs to be written in the Fasta format, and the header line of each sequence should contain the name of the structure file as the first field, and the chain in the last field. Here is an example:

>1FUJ.pdb|PDBID|CHAIN|A
ivggheaqphsrpymaslqm---rgnpgshfcggtlihpsfvltaahclr
dipqrlvn---vvlgahnvrtqeptqqhfsvaqvfln-nydaenklndil
liqlsspanlsasvatvqlpqqdqpvphgtqclamgwgrvgahdppaqvl
qelnvtvvt---------ffcrphnict-fvprrk-a-gicfgdsggpli
cdgiiqgidsfviwgcatrlfpdfftrvalyvdwirstlr
>1PPF.pdb|PDBID|CHAIN|E
ivggrrarphawpfmvslql---rg---ghfcgatliapnfvmsaahcva
nvn---vravrvvlgahnlsrreptrqvfavqrifen-gydpvnllndiv
ilqlngsatinanvqvaqlpaqgrrlgngvqclamgwgllgrnrgiasvl
qelnvtvvt---------slcrrsnvct-lvrg-r-qagvcfgdsgsplv
cnglihgiasfvrggcasglypdafapvaqfvnwidsiiq

Sequences in the Fasta file and PDB files need to be identical, and the sequences in the alignment need to be of the same length (including gaps, i.e. "-"). Only one chain can be compared at a time (as per the alignment), so the chain field must not be empty. The size limit of the submission here is 10000 amino acids per structure, for up to 20 structures.

A quick and easy way to generate a structural alignment of multiple structures is through PDBeFold (Krissinel & Henrick, Acta crys. Section D bio. crys., 2004). You still require editing the output Fasta file according to the format described above.

Examples

Single Analysis

We show the example of analysis done on the bacterial pentameric proton-gated ion channel (GLIC). In the overlap analysis, we used the structure in one of the locally-closed states (PDB 3tlu, cyan, left) and open state (PDB 3eam, magenta, right). We first submitted 3tlu to the Single Analysis submission tab, where WEBnm@ proceeded to calculated its normal modes and return the default analysis. 3eam was then submitted in the Overlap Analysis field and running this produced an animation of the transition from one state to the other, and the two plots characterising this (below).

overlapcum_overlap

The y-axis represents the dot product of the vectors from the mode and the coordinates of the second structure, and in the case of the cumulative plot, this score sums to one. The x-axis represents the mode number. In the overlap plot (left), the peaks represent the modes that contribute significantly to this conformational change, the highest of which is the 11th mode. Since the first 6 modes are ignored as they represent rigid-body rotational/translational movements, the first point starts at mode 7. The mode index for the given scores are also given the in the raw data file (text-based with 'DAT' extension). The cumulative overlap (right) shows the number of modes that give the most coverage.

Comparative Analysis

(Input files used: Fasta file Structure 1 Structure 2 Structure 3 Structure 4)

The single subunits of 3eam, 3tlu were aligned with homologues ELIC (2vl0) and GluCl (3rhw). The output includes the normalized fluctuations profile (left) and the comparative overlap heatmap (right) which uses the Bhattacharya Coefficient (BC) scoring.

fluctuations
cum_overlap

In the normalized fluctuations profile, the structures are represented by 2vl0 (red), 3tlu (green), 3eam (cyan) and 3rhw (purple). The x-axis is the alignment index according to sequence in the proteins while the y-axis represents the score, which is summed, averaged and weighted by each modes' contribution for each position. The comparative overlap heatmap is represented such that each structure pairs with each other, including itself. This gives a diagonal which gives the maximum score of 1.0, while the others are lower, ranging between 0.7 and 0.8. This shows that despite these structures having sequence identities of between 22-32%, their dynamics are much more similar.

Other example inputs for comparative analysis:

P-type ATPases family : P-type files (in zip archive)
20 TIM Barrel proteins : TIMs files (in zip archive)