I'd like to use the loop prediction method that Rich Friesner described in a recent presentation in which he showed extremely accurate prediction of long loops (14-20 residues). Can I access this method using Prime?

Yes, this loop prediction method is available from Suite 2011 (Prime v3.0) on, but not from the graphical interface. Use the Prime Refinement panel to set up your loop prediction job, write out the input file, modify the input file as described below, and submit the job from the command line using the following command:

$SCHRODINGER/multirefine <input file>.inp -HOST

ORIGINAL INPUT FILE (written from the GUI for Ultra-extended loop sampling)		NEW INPUT FILE
STRUCT_FILE	<path>/<input structure>.mae	STRUCT_FILE	<path>/<input structure>.mae
JOB_TYPE	REFINE	JOB_TYPE	REFINE
PRIME_TYPE	4	PRIME_TYPE	4
LOOP_0_RES_0	A:26	LOOP_0_RES_0	A:26
LOOP_0_RES_1	A:40	LOOP_0_RES_1	A:40
MAX_JOBS	0	MAX_JOBS	0
MIN_OVERLAP	0.70	MIN_OVERLAP	0.70
USE_RANDOM_SEED	no	USE_RANDOM_SEED	no
SEED	0	SEED	0
EXT_DIEL	80.00	EXT_DIEL	80.00
USE_MEMBRANE	no	USE_MEMBRANE	no
HOST	<host name>	HOST	<host name>
USE_CRYSTAL_SYMMETRY	no	USE_CRYSTAL_SYMMETRY	yes
RES_SPHERE	7.50
		OFAC_INIT_STAGE	0.45/0.50/0.55/0.60/0.65
		THREADS	10
		NUM_FIXED_STAGE	10
		NUM_OUTPUT_STRUCT	10

Here is an example input file:

STRUCT_FILE   <path>/<input structure>.mae
JOB_TYPE   REFINE
PRIME_TYPE   4
LOOP_0_RES_0   A:26
LOOP_0_RES_1   A:40
MAX_JOBS   0
MIN_OVERLAP   0.60
USE_RANDOM_SEED   no
SEED   0
EXT_DIEL   80.00
USE_MEMBRANE   no
HOST   <host name>
USE_CRYSTAL_SYMMETRY   yes
OFAC_INIT_STAGE 0.45/0.50/0.55/0.60/0.65
THREADS   10
NUM_FIXED_STAGE   10
NUM_OUTPUT_STRUCT   10


Note, make sure to edit the STRUCT_FILE, LOOP_0_RES_0, LOOP_0_RES_1, and HOST lines.

Keywords: Prime, long loop prediction

Back to Search Results