-
Notifications
You must be signed in to change notification settings - Fork 6
/
INSTALL
37 lines (23 loc) · 1.48 KB
/
INSTALL
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
This is GREMLIN v2.01 - a version that learns a Markov Random Field from an input alignment with parameters optimized for accuracy and speed of contact prediction.
You have two options to run gremlin code
1) As a matlab function:
a) compile the mex executables for your system architecture as follows:
within matlab, go to the gremlin dir and type
>mexAll
b) add gremlin to your matlabpath. One way to do this is to go to the gremlin src dir in matlab and then type
>addpath(genpath(pwd));
or,
2) As a compiled executable:
You'll need to compile on a m/c that has a matlab license (with the statistics toolbox) which has the same architecture as the machine you plan
to run on. If you are running on a linux x86-64 machine, you might be able to use the precompiled executables bundled
with this tar-ball. You'll also need the matlab component runtime (MCR) installed on the machine you run the code on
(this can be freely downloaded from the mathworks website without having a MATLAB license).
To compile:
go to the gremlin directory
type on the unix shell (expects mcc to be on your path)
$bash make.sh
This should create an executable called gremlin end an wrapper shell script called run_gremlin.sh
You may use run_gremlin.sh just as you call gremlin but with one additional parameter at the beginning: the path to the MCR
Note:
This version of gremlin uses pdist which is included with the statistics toolkit.
We are working on a providing an alternate which does not need this toolkit.