Download, install, and run

Curiosity: Who are you?
SISA: My name is SISA. I help optmizing experimental phases for a few strongest reflections; you can use my output mtz file to perform density modification.
Curiosity: How do I do that?
SISA: Simple. Just follow these steps.
* You need Phenix to run sisa.

1.Download sisa

2.Unzip the downloaded file to your local folder. From terminal window, enter the unzipped folder.

3.You can try sisa with the sample data by running this script.
phenix.python sisa.py project_name=genev
run_name=desy_iod_pk_try01
hklin=data/data_gene5c2_phib.mtz
hklout=result_phis.mtz
index_param=1,100#101,200#201,300#301,400#401,500
hklbenchin=data/data_gene5c2_phic.mtz column_phic=PHIC flag_autobuild_on=1 file_name_seq=data/genev.fasta file_name_ha=None file_name_model=None semet=1
Curiosity: What are these parameters for?
SISA: Thank you for asking.
Parameters in bold are required parameters. You need to tell me all of them.
project_name: This is your project name. I'll create a folder under my directory {sisa}/projects/{your project name}. I'll stored the output files including the log files here.
run_name: I need your run name (just in case you'd like to try many runs). I'll place a folder with run_name under your project directory.
hklin: This is your experimental phase mtz file. It should have these following columns: FP,SIGFP,PHIB,FOM,HLA,HLB,HLC,HLD. If your mtz file contains different column names, please help me specify them in the same order. You can do this by telling me that columns=FP,SIGFP,PHIB,FOM,HLA,HLB,HLC,HLD.
hklout: Your output mtz file name. I'll output the mtz file that contains the opimized phases for your selected strongest reflections and your original centroid phases for the rest of the reflections. You'll find this file under your {project_name}/{run_name}/ directory.
index_param: You can tell me how many reflections and cycles to run. The index_param=1,100#101,200#201,300#301,400#401,500 means running 5 cycles; first cycle starts with 1st - 100th strongest reflections, second cycle with 101st - 200th strongest reflections, and so on. The index_param=1,500 means running only one cycle for the 1st - 500th strongest reflections.
That's all. Now you're ready to run.
Curiosity: How about other parameters?
SISA: Oh. In case you'd like to perform bechmarking test, you can specify hklbenchin parameter (as mtz file). It should contain phases calculated from the solved structure. I'll assume column name for these calculated phases as PHIC; unless you specify something else for example, column_phic=PHICalc. If I have the hklbenchin, I'll calculate the map correlations (Read, 1986; Lunin & Woolfson, 1993) and output them in the log file.
Curiosity: How do I proceed after running the program?
SISA: You can take your output mtz file for density modification and model building. I can also perform a quick density modification using Phenix.autobuild for you. Please set these parameters:
flag_autobuild=1
file_name_seq={your sequence file}
file_name_ha={your heavy atom or anomalous scatterer file}
file_name_model={your starting model (pdb file) in case you already have one}
semet=1, if you'd like to build the anomalous scatterers as selenium methionine.
I'd recommend this process (running quick Phenix.autobuild here). As soon as I have the density-modified maps, I can tell you about the quality of these new maps. You can compare them with the density-modified maps generated from the original experimental phases to see if there are some improvements.
Please note that I also have other optional parameters to try out. I was born in to the genetic algorihm family; some other optmizing parameters are available. Please contact me for more information.
Curiosity: Okay. Thanks. Bye for now.
SISA: Good bye.

My email

  • Read, R.J. (1986). Acta Cryst. A42, 140-149.
  • Lunin, V. Y. & Woolfson, M. M. (1993). Acta Cryst. D49, 530-533.