- Posts: 24
Atomistic collagen to CG issues
- acnash
-
Topic Author
- Offline
- Junior Boarder
Less
More
7 years 5 months ago #6272
by acnash
Atomistic collagen to CG issues was created by acnash
Dear all, I hope you can help. It's been many years since I lasted used MARTINI (probably 2011) and I am very excited to give it a go with a collagen fibril. My group has been working on full sized all-atom collagen proteins (approx 3154 amino acids), but it's time we switched to CG given the size of the structure.
I am using the very latest martinize distribution for python 2.6. The input pdb file is an all-atom human type-1 collagen molecule (three polypeptide strands to form the right-handed superhelical structure). Thankfully, MARTINI now supports hydroxyproline, however, I have had to implement into the code hydroxylysine (I just copied all of lysine and adjusted the polarity of the central bead accordingly).
Unfortunately, when running the script using:
python2.6 ../MARTINI/martinize_python2-6.py -f test.pdb -o test_cg.top -x test_cg.pdb -collagen -v
The output is no longer of three chains delimited by two "TER" identifiers, but it has been broken up into many chains:
INFO Input structure is a PDB file.
DEBUG Splitting chain in 35 chains
INFO Found 35 chains:
INFO 1: (Protein), 1350 atoms in 104 residues.
INFO 2: (Unknown), 17 atoms in 1 residues.
INFO 3: (Protein), 10315 atoms in 842 residues.
INFO 4: (Unknown), 17 atoms in 1 residues.
INFO 5: (Protein), 1268 atoms in 99 residues.
INFO 6: (Unknown), 17 atoms in 1 residues.
INFO 7: (Protein), 925 atoms in 73 residues.
INFO 8: (Unknown), 17 atoms in 1 residues.
INFO 9: (Protein), 494 atoms in 35 residues.
....<CONTINUES>
INFO Removing HETATM chain consisting of 1 residues.
INFO Removing HETATM chain consisting of 1 residues.
INFO Removing HETATM chain consisting of 1 residues.
INFO Total size of the system: 3133 residues.
INFO Writing coarse grained structure.
....<CONTINUES>
DEBUG FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
DEBUG
INFO Created coarsegrained topology
DEBUG FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF......
.....<CONTINUES It repeats with each chain, the amino acid content and the secondary structure type, for all chains. >
INFO Created coarsegrained topology
DEBUG FFFFFFFF
DEBUG
INFO Created coarsegrained topology
INFO Written 16 ITP files
INFO Output contains 18 molecules:
INFO 1-> Protein (chain )
INFO 2-> Protein (chain )
INFO 3-> Protein (chain )
INFO 4-> Protein (chain )
INFO 5-> Protein (chain )
INFO 6-> Protein (chain )
INFO 7-> Protein (chain )
INFO 8-> Protein (chain )
INFO 9-> Protein (chain )
INFO 10-> Protein (chain )
INFO 11-> Protein (chain )
INFO 12-> Protein (chain )
INFO 13-> Protein (chain )
INFO 14-> Protein (chain )
INFO 15-> Protein (chain )
INFO 16-> Protein (chain )
INFO 17-> Protein (chain )
INFO 18-> Protein (chain )
INFO Written topology files
INFO Note: Cysteine bonds are 0.24 nm constraints, instead of the published 0.39nm/5000kJ/mol.
<END>
First of all, only one .itp was created, and secondly the output CG pdb file is now delimited into more than the original three chains for all 3154 residues.
I tried removing the "-collagen" flag, but it made no difference. Has anyone any thoughts of why the output topology isn't three complete collagen polypeptide chains as per the original input pdb file?
Many thanks
Anthony
I am using the very latest martinize distribution for python 2.6. The input pdb file is an all-atom human type-1 collagen molecule (three polypeptide strands to form the right-handed superhelical structure). Thankfully, MARTINI now supports hydroxyproline, however, I have had to implement into the code hydroxylysine (I just copied all of lysine and adjusted the polarity of the central bead accordingly).
Unfortunately, when running the script using:
python2.6 ../MARTINI/martinize_python2-6.py -f test.pdb -o test_cg.top -x test_cg.pdb -collagen -v
The output is no longer of three chains delimited by two "TER" identifiers, but it has been broken up into many chains:
INFO Input structure is a PDB file.
DEBUG Splitting chain in 35 chains
INFO Found 35 chains:
INFO 1: (Protein), 1350 atoms in 104 residues.
INFO 2: (Unknown), 17 atoms in 1 residues.
INFO 3: (Protein), 10315 atoms in 842 residues.
INFO 4: (Unknown), 17 atoms in 1 residues.
INFO 5: (Protein), 1268 atoms in 99 residues.
INFO 6: (Unknown), 17 atoms in 1 residues.
INFO 7: (Protein), 925 atoms in 73 residues.
INFO 8: (Unknown), 17 atoms in 1 residues.
INFO 9: (Protein), 494 atoms in 35 residues.
....<CONTINUES>
INFO Removing HETATM chain consisting of 1 residues.
INFO Removing HETATM chain consisting of 1 residues.
INFO Removing HETATM chain consisting of 1 residues.
INFO Total size of the system: 3133 residues.
INFO Writing coarse grained structure.
....<CONTINUES>
DEBUG FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF
DEBUG
INFO Created coarsegrained topology
DEBUG FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF......
.....<CONTINUES It repeats with each chain, the amino acid content and the secondary structure type, for all chains. >
INFO Created coarsegrained topology
DEBUG FFFFFFFF
DEBUG
INFO Created coarsegrained topology
INFO Written 16 ITP files
INFO Output contains 18 molecules:
INFO 1-> Protein (chain )
INFO 2-> Protein (chain )
INFO 3-> Protein (chain )
INFO 4-> Protein (chain )
INFO 5-> Protein (chain )
INFO 6-> Protein (chain )
INFO 7-> Protein (chain )
INFO 8-> Protein (chain )
INFO 9-> Protein (chain )
INFO 10-> Protein (chain )
INFO 11-> Protein (chain )
INFO 12-> Protein (chain )
INFO 13-> Protein (chain )
INFO 14-> Protein (chain )
INFO 15-> Protein (chain )
INFO 16-> Protein (chain )
INFO 17-> Protein (chain )
INFO 18-> Protein (chain )
INFO Written topology files
INFO Note: Cysteine bonds are 0.24 nm constraints, instead of the published 0.39nm/5000kJ/mol.
<END>
First of all, only one .itp was created, and secondly the output CG pdb file is now delimited into more than the original three chains for all 3154 residues.
I tried removing the "-collagen" flag, but it made no difference. Has anyone any thoughts of why the output topology isn't three complete collagen polypeptide chains as per the original input pdb file?
Many thanks
Anthony
Please Log in or Create an account to join the conversation.
- peterkroon
-
- Offline
- Gold Boarder
Less
More
- Posts: 210
7 years 5 months ago #6308
by peterkroon
Replied by peterkroon on topic Atomistic collagen to CG issues
My guess is that it doesn't recognize your hydroxylysine as protein chain (INFO 2: (Unknown), 17 atoms in 1 residues.) and cuts the chains there. Does that number add up to the number of chains you get?
Please Log in or Create an account to join the conversation.
- acnash
-
Topic Author
- Offline
- Junior Boarder
Less
More
- Posts: 24
7 years 5 months ago #6312
by acnash
Replied by acnash on topic Atomistic collagen to CG issues
I'll be a bit nieve and say, yes it does, simply because I'm away from my machine for the next couple of days and from memory the number of diving chains sounds similar to the number of hydroxylysine. I replicated every entry of lysine in the script and replaced it with a new name and bead polarity, so I can't possibly why it couldn't recognise LYZ. I'll keep hacking through the script, if I find out what the problem was I'll post it here.
Please Log in or Create an account to join the conversation.
- acnash
-
Topic Author
- Offline
- Junior Boarder
Less
More
- Posts: 24
7 years 5 months ago #6317
by acnash
Replied by acnash on topic Atomistic collagen to CG issues
Problem resolved. It turns out that hydroxylysine was not the problem but rather it was histidine. The original MD all-atom structure was derived from an AMBER forcefield, which, like many other forcefields, has a number of variation to histidine according to bond order. MARTINI comes with HIS and HIP, whereas AMBER has HIE and HID.
Thanks for the suggestion of looking at atomtype recognition, it turned out to be the case, but just for a different residue.
Thanks for the suggestion of looking at atomtype recognition, it turned out to be the case, but just for a different residue.
Please Log in or Create an account to join the conversation.
Time to create page: 0.091 seconds