Users can upload their own structures in PDB format or choose a query structure already submitted in the PDB to test their own potentials and the like. For this, the input can be just the 4 letter PDB code. Individual interfaces can also be specified.

Using this option, the user can upload their own potentials for scoring interfaces. There are two types of potentials:

1) Residue pairing preferences (Scoring Matrix)

2) Atomic Propensity

Users can upload up to three scoring matrices and three atomic propensity matrices, one for each interaction type. It is not compulsory to upload all 6 matrices and in cases where a user uploads fewer matrices, PIZSA will use the default matrices for rest of the potentials.

All custom potential files are required to be comma separated files (CSV). The format for Scoring Matrix and Atomic Propensity files are different and are described below:

Scoring Matrix file format:

[Res1],[Res2],[Score]

where:

Res1 - Residue 1

Res2 - Residue 2

Score - Residue pairing score (Real Number)

Example:

GLU,HIS,0.864

ILE,SER,-2.307

ASP,MET,-1.247

MET,TRP,1.112

ALA,PRO,-2.529

GLN,TRP,0.866

GLY,PRO,-4.183

GLN,PRO,-0.570

ASP,ASP,-0.201

GLY,LYS,-4.183

Note:

1) Files should have 1 to 400 lines (rows)

2) Files should have 3 comma(,) separated fields (columns)

3) Reverse residue pairs have the same score as the original residue pair, unless user specifies a different score

→ eg. 1. Symmetric scores

Input:

ASP,MET,-1.247

Scores:

ASP-MET = -1.247

MET-ASP = -1.247

→ eg. 2. Asymmetric scores

Input:

ASP,MET,-1.247

MET,ASP,2.021

Scores:

ASP-MET = -1.247

MET-ASP = 2.021

4) Unspecified residue pairs are automatically assigned a score of 0

Atomic Propensity file format:

[Res1],[Res2],[AtomCount],[Score]

where:

Res1 - Residue 1

Res2 - Residue 2

AtomCount - Number of contact atoms (Integer)

Score - Residue pairing score (Real Number)

Example:

ARG,GLU,2,0.46

ARG,GLU,3,0.42

ARG,GLU,4,0.93

ARG,GLU,5,1.10

ARG,GLU,6,3.00

ARG,GLU,7,3.00

ARG,GLU,8,4.00

ARG,GLU,9,4.00

ARG,GLU,10,1.00

ARG,GLU,11,0.00

Note:

1) Files should have 1 to 400 lines (rows)

2) Files should have 4 comma(,) seperated fields (columns)

3) Reverse residue pairs have the same score as the original residue pair, unless user specifies a different score

→ eg. 1. Symmetric scores

Input:

ARG,GLU,7,3.00

Scores:

ARG-GLU[7] = 3.00

GLU-ARG[7] = 3.00

→ eg. 2. Asymmetric scores

Input:

ARG,GLU,7,3.00

GLU,ARG,7,1.50

Scores:

ARG-GLU[7] = 3.00

GLU-ARG[7] = 1.50

4) Unspecified residue pairs are automatically assigned a score of 0

5) AtomCount can have integer values from 2 to 28

6) Scores can have real number values greater than or equal to 0

Note: MC-SC matrices should ideally be asymmetric since the contribution depends on what residue interacts via the MC atoms and what residue interacts via their SC atoms. In MC-SC matrices Res1 interacts via MC atoms and Res2 interacts via SC atoms.

For details, a sample file is provided here: Custom Potential File

Two amino acid residues are defined as interacting if any relevant atom of residue A was within any relevant atom of residue B. The user can specify different distance thresholds for defining interface residue contacts and choose between 4 Å, 6 Å and 8 Å.

Note: A smaller distance threshold will provide a more stringent definition of the interface whereas a larger threshold might be helpful in the case of a loosely binding complex.

This parameter is useful when certain types of interactions are dominant in the protein complex. The users can choose to isolate interactions between certain atoms and use those to define the interface. Choices for this parameter include side chain-side chain interactions, main chain-main chain interactions and main chain-side chain interactions.

Our server makes use of 48 different statistical potentials that have been constructed using a combination of different interaction distance thresholds and type of interacting atoms. Based on the input parameters given by the user, a particular statistical potential will be used and the corresponding Z-score threshold be employed to make the prediction.

Amino acid residues on the interface are mutated in-silico and the interfaces rescored to estimate the contribution of individual residues towards the interface. The user is presented with three options in this regard.

1. Interface Alanine Scanning: This option will mutate all the interface residues into 'Alanine'.
2. Interface Saturated Mutagenesis: This option will perform saturated mutagenesis for all the interface residues.
3. Manual mutation specification: This option allow to manually specify the residues to be mutated. Further, user can mutate specific interface residue to other residue(s), to alanine or do a saturated mutagenesis for the selected residue.

The table below specifies how user can specify mutations to be done. A sample file is also provided to illustrate how different type of mutations can be specified by the user.

---

Mutate a given list of residues to every other residue, a sample mutagenesis file . Each line of the file has two tab separated fields
Mutation Type	Field 1	Field 2
Mutating the residue to Alanine (mutate the residue number 177 (valine) in chain 'B' to Alanine)	177:VAL:B	ALA
Mutating a residue to every other residue (mutate the residue number 20 (histidine) in chain 'A' to all the other residues, Saturated mutagenesis)	20:HIS:A	*
Mutating the residue to specific residue (mutate residue number 27 (gultamate) in chian 'A' to lysine)	27:GLU:A	LYS
Mutating the residue to a list of residue (mutate residue number 27 (gultamate) in chian 'A' to lysine and valine)	27:GLU:A	LYS,VAL

Description of Output The most important residues are listed in a ranked order list. The last column 'Score-Diff' shows the magnitude of change in scores upon mutagenesis. The positive values of this column imply that the mutation destabilizes the interface whereas negative values imply that the mutation stabilizes it.

The jobname is a name given by the user for the job. The job name is useful in distinguishing between different jobs.

Jobname is also reflected in the name of the output file created by the program.

If the user provide an email address then an email notification will be sent to inform the user about the job completion or unexpected termination of the job.

It is advised that a valid and active email address must be provided to receive notifications about the job status.