Pharmer Activity
Status: Beta
Brought to you by:
dkoes
Activity for Pharmer
-
David Koes posted a comment on discussion Help
pharmer doesn't do that. Check out PharmaGist: https://pmc.ncbi.nlm.nih.gov/articles/PMC2447755/
-
JONAS_Li posted a comment on discussion Help
I have a cluster of 50 ligands that have similar structures, and I want to use them to build a consensus pharmacophore without target(protein).
-
Here is the code from my version of load_query which seems to work: def get_neighbors(points, point, num_neighbors): ''' Naive method for finding the nearest neighbors of a point on a sphere of points Arguments: points -- list of points (numpy arrays) point -- point of interest num_neighbors -- number of neighbors to return Returns: neighbors -- list of neighboring points ''' distances = [ [pt, np.linalg.norm(point - pt)] for pt in points] pts = np.stack(list(map(lambda x: x[0], sorted(distances,...
-
How to fix the folowwing errors: >load_query ph_vx548.json Aromatic 130.5346666666667 138.4101666666667 131.3473333333333 enabled = True rad = 1.100000023841858 Traceback (most recent call last): File "D:\pymol\lib\site-packages\pmg_tk\startup\load_query_python3.py", line 526, in load_query spherelist += make_mesh(x,y,z, rad, color = colors) File "D:\pymol\lib\site-packages\pmg_tk\startup\load_query_python3.py", line 304, in make_mesh neighbors = get_neighbors(points, point, num_neighbors) File "D:\pymol\lib\site-packages\pmg_tk\startup\load_query_python3.py",...
-
JonasLi posted a comment on discussion Help
Do you mean I need to compile and install the pharmer with a newer openbabel(like 3.1.1)? Because I used to directly use ./pharmer.static
-
Based on the N with 4 bonds it looks like you need a newer version of openbabel (e.g., current git version).
-
I meet the same question for wrong order of N atom and maybe I need install old version openbabel. But I cannot find 2.3.2 or former version...
-
Hi Dr.Koes, I have installed the load_query script in Pymol Plugin, but when I use load_query command to show my ph.json, it has numpy errors: ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 2 dimensions. The detected shape was (100, 2) + inhomogeneous part. My json file is as follows, whose vector information has been deleted: { "points" : [ { "name" : "HydrogenDonor", "radius" : 0.50, "requirement" : "required", "size" : 1, "x" : -1.0010,...
-
I'm not sure I can elaborate beyond the error message much. You have molecules next to each other in the file that have different chemistries but have the same name. This isn't allowed since names are used to uniquely identify molecules.
-
My server is Ubuntu, I download a molecule database sdf file "T001.sdf" to do pharmacophore screening, and I runs the "./pharmer.static dbcreate -in T001.sdf -dbdir db" and the result is as following..... Warning: Invalid Input. Sequential molecules with the same name should be conformers. Warning: Invalid Input. Sequential molecules with the same name should be conformers. Warning: Invalid Input. Sequential molecules with the same name should be conformers. Creating spatial index... 14492 conformations...
-
Warning: Invalid Input. Sequential molecules with the same name should be conformers. Warning: Invalid Input. Sequential molecules with the same name should be conformers. Warning: Invalid Input. Sequential molecules with the same name should be conformers. Creating spatial index... 14492 conformations 14259 molecules 7708 internal pages Index creation time: 1.91s (3.18s) I want to know why this warning comes up? How to deal with it during database creation?
-
what does this means, why it works
-
This isn't an appropriate forum to raise issues with ZINCPharmer. I suggest you email the person responsible for that website or use Pharmit instead.
-
Srinjana raha posted a comment on discussion Help
Zinc pharmer says ' error connecting to search engine. Please try again later and if problem persists, contact the administrator ' since two days. Tried from several laptops. tried with several different molecules. Both the feature and receptor not loading.
-
Yes, the vector matching is approximate so when you have many of them the accumulated approximations can result in too narrow of a query.
-
David Johnson posted a comment on discussion Help
lig2.sdf was copied from one conformer of my conformer file, as generated by rdkit-scripts/rdconf.py . Removing the vectors got it to run and return the hit I was expecting. Should I be removing the vectorrs?
-
How was lig2.sdf created? What happens if you remove all the vectors from the query?
-
I’m trying to get Pharmer up and running, and I’m not getting the results that I would expect. I used Dr. Koes rdkit-scripts on Github to generate 5k conformers in a SDF file, and used it to build a DB. For a compound to test, I took the very first compound and used it to build a pharmacophore. I had actually output the pharmacophore as an SDF. When I run "pharmer.static dbsearch -dbdir DB -in pharm2.sdf” it returns "Invalid extension for query file: pharm2.sdf”, which is not expected based on the...
-
Hydrogens aren't stored nor output. You can add them back with your favorite cheminformatics toolkit.
-
Himanshu Goel posted a comment on discussion Help
Hi, For pharmacophore screening with million of compounds, I recently observed the difference in the protonation states between the input mol2 files and output sdf files. This is an important issue and need to be addressed. I have attached one example in this regard(mol2 input file and sdf output file). Please let me know how this issue can be resolved or I am missing something in that regard.
-
Well, that line in mimic_pmg_tk.py is import tkinter Do you have this package installed? If you don't need the plugin, you can just run the script and use the command.
-
Raziel Cedillo posted a comment on discussion Help
Hi everyone. I've had some isssues with this plugin for pymol. Pymol gives me the following error message: RuntimeError: tkinter init failed I was wondering if there is any way to solve this error Regards, Te all message is: Traceback (most recent call last): File "/Users/raziel/miniconda3/envs/Chem3/lib/python3.6/site-packages/pmg_tk/startup/load_query_python3.py", line 806, in <lambda> self.menuBar.addmenuitem('Plugin', 'command', 'Pharmacophore Query Loader', label = 'Pharmacophore Query Loader',...
-
Pharmer released /load_query_python3.py
-
-
pharmer is a standalone executable. You will not get this error with pharmer, which is what this discussion board is for, since it is not a web service. Perhaps you want to use pharmit (http://pharmit.csb.pitt.edu/)?
-
Nilesh Bajad posted a comment on discussion Help
Please help with following error " Error connecting to search engine. Please try again later and if the problem persists, contact the administrator"
-
Gautam Bhardwaj posted a comment on discussion Help
Thank you for your response. With regards Gautam Bhardwaj On Mon, 24 Aug, 2020, 6:57 PM David Koes, dkoes@users.sourceforge.net wrote: Unfortunately there isn't any documentation beyond what is provided by the application itself when you run it. There are a lot of similarities to the pharmit interface though: https://sourceforge.net/projects/pharmit/files/ Pharmer User Manual/ Pharmacophore https://sourceforge.net/p/pharmer/discussion/1669176/thread/c3da0e36/?limit=25#f7c0/107c/244a Sent from sourceforge.net...
-
Unfortunately there isn't any documentation beyond what is provided by the application itself when you run it. There are a lot of similarities to the pharmit interface though: https://sourceforge.net/projects/pharmit/files/
-
Is it work on cygwin platform but how to run if possible please hint. I try. Thank you with regards Gautam bhardwaj On Sun, 23 Aug, 2020, 4:53 PM David Koes, dkoes@users.sourceforge.net wrote: I have never tried to build it on Windows. Perhaps if you first install the Linux Subsystem it might work. Pharmer User Manual/ Pharmacophore https://sourceforge.net/p/pharmer/discussion/1669176/thread/c3da0e36/?limit=25#f7c0 Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/pharmer/discussion/1669176/...
-
I have never tried to build it on Windows. Perhaps if you first install the Linux Subsystem it might work.
-
is pharmer run on window and how toinstall .please provide information . its manual also please
-
chris posted a comment on discussion Help
try running this in bash and then running pharmer again export LC_ALL=C if it works add it to your .bashrc file
-
It looks like you'll have to build it from source for it to run on your configuration.
-
Pir Wlan posted a comment on discussion Help
Thank you for getting back! I had a look at pharmit. I would like to run it locally, as I have thousands of pdb files to screen. When I run the executable, I get the same error message as described above when trying to run the static pharmitserver file.
-
Try using pharmit (https://sourceforge.net/projects/pharmit/files/) I don't keep the pharmer executable up to date and its functionality is a subset of pharmit.
-
Dear David, Due to my lack of Linux skills, I am trying to run the pharmer.static executable, instead of compiling pharmer myself. I tested this on several Ubuntu systems (mostly vBoxes) and always get the following error message: pharmer.static: loadlocale.c:129: _nl_intern_locale_data: Assertion `cnt < (sizeof (_nl_value_type_LC_TIME) / sizeof (_nl_value_type_LC_TIME[0]))' failed. Aborted (core dumped) I found a workaround, which worked for one Ubuntu system, yet then I got: terminate called after...
-
Sulee posted a comment on discussion Help
Dear Dr. Koes, Oh OK, I get it now. I was expecting more features when I used the receptor as input too. Thanks so much for clarifying.
-
That should work. The interaction pharmacophore is a subset of the ligand-only pharmacophore. It consists of those ligand features with matching features on the receptor. What makes you think the output is not based on the interaction?
-
I'm trying to generate an interaction pharmacophore using a ligand and its binding site: ./pharmitserver --cmd pharma -in 4pxm_1_ligand.sdf -receptor 4pxm_1_site.mol2 -out test_pharma.json I used this command, but the pharmacophore that gets generated seems to be only based on the ligand. Could somebody please tell me how to get an interaction pharmacophore correctly? Thank you!
-
Your previous question was about comparing two pharmacophores. That is not what you are doing here. You are comparing the results of two pharmacophore searches. The RMSD in this context is the RMSD between the pharmacophore features of the hits and the matching features of the query. There is a maximum possible RMSD given the radii of the query tolerance spheres (e.g., if the radii are all 1, is is impossible to have an RMSD greater than 1.0). If max-rmsd is larger than this it will (correctly) have...
-
Dear Dr. Koes, Thanks for your answer. I m using pharmer binary to screen through a database. ./pharmer.static -cmd dbsearch -dbdir DB_test -max-rmsd=15 -in $file -extra-info -out out.sdf Example outputs: Query 1c5n_1.json NumResults: 0 Time: 0.01 Query 1ai0_3.json 0,0.3161377,430,7,,1,65536 1,0.35341495,341,12,,3,196608 2,0.35341495,340,12,,5,327680 3,0.35341531,341,12,,3,196608 4,0.35341531,340,12,,5,327680 5,0.56659698,430,7,,1,65536 NumResults: 6 No matter how much I increase the RMSD cutoff,...
-
You would have to define a similarity function between pharmacophores and then implement it. I know of no standard pharmacophore similarity. For rmsd the pharmacophores need to have identical features and you need a mapping between them.
-
How can I get the similarity/RMSD between two json pharmacophores I generated using pharmer? Thank you!
-
Jiafeng Liao posted a comment on discussion Help
Dear David, Apologize for the late reply. I hvae been off my computer for two days. The input molecules have correct bond orders. I attached the cases so that you can reproduce the bug. If there is any other information needed, please feel free to contact me. Kind regards Jiafeng
-
Can you make a reduced test case for openbabel (e.g. using babel) that shows the problem and we can see if we can get a fix into openbabel 3.0? I assume you have verified that your input molecule doesn't have the incorrect bond orders to start?
-
I built a pharmer database containing 10 conformers of molecules. Then did search on the database. The returned .sdf using [-out out.sdf] during dbsearch consists of molecules with incorrect bond order compared to the original sdf files used to create the database. I have tried using openbabel 2.3.2 and 2.4.1, both have the same issue.
-
I built a pharmer database containing 10 conformers of molecules. Then did search on the database. The returned .sdf using [-out out.sdf] during dbsearch consists of molecules with incorrect bond order compared to the original sdf files used to create the database. I have tried using openbabel 2.3.2 and 2.4.1, both have the save issue.
-
What exactly did you try?
-
Dear David, In order to get aliphtic amine structures from our data, we construct a query with positive and donor features, we could get desired results on zincpharmer website. However, when we use chemdiv 3d conformation library made by ourselves, we only could get aromatic-heterocycle compounds. I attached the results below, could you have a look and see what is the cause?
-
Dear David, I just tried 2.3.2 release and the bond order is still incorrect. Kind regards Jiafeng
-
There is no tagged release with that number. It is the development (master) branch at that date. I don't know when that bug was introduced, but I would assume the 2.3.2 release is okay.
-
Hi, David, I still cannot find that specific branch. Is the bug only existing in newer version than 2.3.90?
-
The .90 suffix means it was from the development branch.
-
Hi, I searched this specific version and found there is no 2.3.90 version on sourceforge or github. Only had 2.3.2 instead. Can you remember where the source was from? Kind regards Jiafeng
-
Thank you so much! I will try that out
-
Actually, I can check: Open Babel 2.3.90 -- Feb 16 2015 -- 13:27:43
-
I'm not sure and don't have a good way to check, but probably whatever was in the Ubuntu 16.04 distribution?
-
Thank you David. May I ask which version you used in ZINCPharmer?
-
ZINCPharmer is using an older version of OpenBabel that doesn't have this bug.
-
Yes, that is openbabel getting the bond orders wrong.
-
Dear David, Following the previous problem. As I found the result from ZincPharmer did not have the 'level of bonds' issues. Is ZincPharmer using the same result generating mechanism as what Pharmer -out out.sdf does? Kind regards Jiafeng
-
Hi, I attached the problem and pointed it out with red marker.
-
Hi, I attached the problem and pointed it out with red marker.
-
If the concern is the change in bond orders, this is a result of openbabel processing the molecule. You will get different results with different versions of openbabel.
-
Dear David, I found the output .sdf after dbsearch is sometimes different from the original input when creating the database as attached. Not sure why it is happening. Any idea or suggestion? Sincere regards Jiafeng
-
Hi, Sure. Here are the things I've done: 1. download: PyMol for python2 2. Use load_query on PyMol for query visualisation 3. Use Pharmer to perform search on database and add "-out output_name.sdf" on command. PS: the filename must have .sdf as suffix to save the search result. 4. Put .sdf file to PyMol and it's done
-
Hi, Sure. Here are the things I've done: 1. download: PyMol for python2 2. Use load_query on PyMol for query visualisation 3. Use Pharmer to perform search on database and add "-out output_name.sdf" on command. PS: the filename must have .sdf as suffix to save the serch result. 4. Put .sdf file to PyMol and it's done
-
Thank you David, for all the helpful advises and the great tools!
-
Also, uploading to pharmit and taking a screen shot is always an option.
-
Sorry, I didn't realize my email response didn't get cc'ed here: the load_query plugin for pymol can read pharmer/pharmit json files: https://sourceforge.net/projects/pharmer/files/load_query_1_1.tar.gz/download
-
ab posted a comment on discussion Help
Dear Jiafeng, I am dealing with a similar representation as well. Could you tell me how you achieved it? Kind regards
-
[PROBLEM SOLVED] By -output .sdf on dbsearch command Dear David, I am currenly working Pharmer on my own database, trying to visualise the query and the search result like what you shown on the paper :https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3124593/figure/F5/. How I can achieve that? Sincere regards
-
Dear David, I am currenly working Pharmer on my own database, trying to visualise the query and the search result like what you shown on the paper :https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3124593/figure/F5/. How I can achieve that? Sincere regards
-
Thank you. It works now.
-
What makes you say it is deprecated? It looks like it is still being developed to me: https://github.com/gperftools/gperftools On 6/10/19 8:34 AM, Jiafeng Liao wrote: Dear David, During the installation, I found these dependencies are deprecated: tcmalloc (google performance tools) http://code.google.com/p/google-perftools/ Is there any alternatives? svn URL invalid https://sourceforge.net/p/pharmer/discussion/1669176/thread/dd9e495e75/?limit=25#d980 Sent from sourceforge.net because you indicated...
-
Dear David, During the installation, I found these dependencies are deprecated: tcmalloc (google performance tools) http://code.google.com/p/google-perftools/ Is there any alternatives?
-
Appreciate. I will try it out.
-
Sorry - responded pre-coffee. There is no git for pharmer. -Dave On Jun 6, 2019, at 08:57, David Koes dkoes@users.sourceforge.net wrote: Also, use git: git.code.sf.net/p/smina/code On Jun 6, 2019, at 8:09 AM, David Koes dkoes@users.sourceforge.net wrote: Remove “trunk” from the tail of the URL. On Jun 6, 2019, at 5:01 AM, Jiafeng Liao jiafeng@users.sourceforge.net jiafeng@users.sourceforge.net wrote: Hi, I tried svn checkout https://svn.code.sf.net/p/pharmer/code/trunk https://svn.code.sf.net/p/pharmer/code/trunk...
-
Also, use git: git.code.sf.net/p/smina/code On Jun 6, 2019, at 8:09 AM, David Koes dkoes@users.sourceforge.net wrote: Remove “trunk” from the tail of the URL. On Jun 6, 2019, at 5:01 AM, Jiafeng Liao jiafeng@users.sourceforge.net jiafeng@users.sourceforge.net wrote: Hi, I tried svn checkout https://svn.code.sf.net/p/pharmer/code/trunk https://svn.code.sf.net/p/pharmer/code/trunk https://svn.code.sf.net/p/pharmer/code/trunk https://svn.code.sf.net/p/pharmer/code/trunk pharmer-code on both eclipes...
-
Remove “trunk” from the tail of the URL. On Jun 6, 2019, at 5:01 AM, Jiafeng Liao jiafeng@users.sourceforge.net wrote: Hi, I tried svn checkout https://svn.code.sf.net/p/pharmer/code/trunk https://svn.code.sf.net/p/pharmer/code/trunk pharmer-code on both eclipes and command tool line directly, it shows: svn: E170000: URL 'https://svn.code.sf.net/p/pharmer/code/trunk' doesn't exist could you have a check to see if the URL is down? svn URL invalid https://sourceforge.net/p/pharmer/discussion/1669176/thread/dd9e495e75/?limit=25#8beb...
-
Hi, I tried svn checkout https://svn.code.sf.net/p/pharmer/code/trunk pharmer-code on both eclipes and command tool line directly, it shows: svn: E170000: URL 'https://svn.code.sf.net/p/pharmer/code/trunk' doesn't exist could you have a check to see if the URL is down?
-
What do you want this score to mean? RMSD is an unbounded metric (molecules can be arbitrarily far away). You could apply a squashing function like a sigmoid but I question the utility.
-
Hello David, I would like to get a score ranged between 0 and 1 (like Tanimoto Coefficient and other measures do). What is the easiest way to achieve so? Is it possible to do it by using any parameter? Thanks
-
What do you mean by nothing happens? You run it and get no output, not even an error message? Which binary did you download and what operating system are you using?
-
Lereze Marais posted a comment on discussion Help
Hi I tried to download pharmer, it downloads but after that nothing happens. do I have to download a additional program?
-
Cenk A Andac modified a comment on discussion Help
-
-
Let me quickly answer, YES. I figured it out..
-
Hello again! I have spent some moment with pharmit, it is quite useful. Is it possible to upload the receptor file from our harddisk rather than typing in PDB ID ? Thanks,
-
You can concatenate the files or provide them all separately on the commandline.
-
pharmit works! Thank you! It worked for one .sdf file though. I have at least 5 more sdf files in the current directory. Do I need to run the command line for each file ? or it is just alright to concatenate all files to run pharmit? Thanks,
-
https://sourceforge.net/projects/pharmit/files/README/
-
Thank you for the kind response. I have looked at the Pharmit Database . It seems quite useful. I will certainly give it a try. On the other hand, I would like to test several conformers (from 1 upto 5/6) of ligands in the database to compare between them.. I still would be interested in those command lines.. Thank you ...
-
Why not use the pre-built ZINC library in Pharmit (http://pharmit.csb.pitt.edu/). What is the advantage of creating your own database?
-
Hi, I have downloaded several database files (in sdf.gz format) off the Zinc15 Database website. I have saved all the sdf.gz files in ./zinc-db directory and gunzip'ped them all therein. I think there are around 3000 sdf files in the database. On the other hand, I generated a query.json file for a ligand that is known to have high binding affinity towards a receptor (using the command line : pharmer pharma -in lig.pdb -out lig.json). I have made a few changes in my lig.json file using the online...
-
It works now. Thank you!
-
Install the simplejson python module
-
Oppss!! Just to correct a typo above, the version of Pymol I use is 1.7.x
-
Hello, I have just installed Pymol 2.1 and tried to install the plugin load_query.py script in pymol. After installation pymol gives me the following error message : line 80 in module import simplejson as json ImportError: No module named simplejson I was woundering if there is any way to troubleshoot the error message. Regards,
-
I'm a bit surprised you are seeing this with 18.04. Maybe something changed in Eigen again. Here is the note on this problem from the readme: Some versions of eigen3 will trigger an assert, OBJECT_ALLOCATED_ON_STACK_IS_TOO_BIG, at compile time. This can be ignored in the latest versions of eigen3 with -DEIGEN_DISABLE_STACK_SIZE_ASSERT If your version hasn't had this patch applied (which is currently true with Ubuntu 14.04), then you will have to modify Eigen/src/Core/DenseStorage.h appropriately...
1 >
