Hi, In addition to all the great suggestions, you can also look into using Rosetta_MR/Phenix and CNS DEN refinement. Can be useful if all fine with your data and space group analysis and primary issue is remote model homology/significant model dissimilarity. Rosetta MR: http://www.nature.com/nature/journal/v473/n7348/abs/nature09964.html CNS DEN: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3322598/
Best, Debanu Das > On Jan 2, 2017, at 12:32 PM, Eleanor Dodson <[email protected]> wrote: > > I would try harder to get an MR solution from your best homologue. > > Things I would do.. > 1] Check data quality - CCP4I2 gives you a good report - Are there ice rings? > strange wilson plot? missing low resolution data? etc etc.. > > 2] is there Non xrystallographic translation? ccould you have the spece group > wrong? > > > 2] have you used chainsaw or sculptor to trim the model? looked carefully at > sequence alignments? how do the known structures align? > > > > You might well be able to find the Fe positions from anomalous scattering if > you have reasonably good data . > Eleanor > >> On 2 January 2017 at 17:13, Claudia Millán Nebot <[email protected]> wrote: >> Dear Madhu, >> >> At the resolution that you mention, which is at the edge of the resolution >> limit for ARCIMBOLDO_LITE, and considering that you have already some >> information about the possible fold, I would suggest you to use our tool >> ARCIMBOLDO_SHREDDER, as it will derive fragments starting from your distant >> homolog template, and refine them against the experimental data, increasing >> the possibilities of convergence of your MR trial. >> >> The current version of CCP4i1 has already an interface to >> ARCIMBOLDO_SHREDDER, which will be updated again in the next days, and you >> can use the spherical mode for your case. >> >> Best wishes, >> >> Claudia Millán >> >> 2017-01-02 14:34 GMT+01:00 Madhu Sudhan <[email protected]>: >>> Dear all, >>> >>> >>> >>> We are trying to solve a protein structure of 37KDa using molecular >>> replacement method. Protein secondary structure indicates that it has >>> 4-Heat repeat (α-helical hair pin) i.e. 16 α-helices. It has about 40% >>> sequence identity with templates in PDB, using which we tried MR, but we >>> are unable to find a solution using MOLREP, PHASER, BALBES, MRBUMP, MORDA >>> etc. >>> >>> This protein has two iron molecules also and data were collected at home >>> source using Cu Ka radiation with 1.5418 A wavelength at 2.5 A resolution. >>> Its highest homology PDB template is also having 16 α-helices with >>> horseshoe shaped structure. >>> >>> >>> >>> We are also trying to use Arcimboldo tool (in the CCP4 7.0) to find the >>> structure solution, but it is taking a long time on our computers (8-core). >>> >>> So I will be grateful if anyone can provide us suggestions >>> >>> 1. Any program which can provide us good working model having only >>> α-helical content. >>> >>> 2. In case we know our model, how can we optimize Arcimboldo tool with our >>> helical model (we can generate from homologous PDB) >>> >>> >>> >>> Thanking you in advance, >>> >>> >>> >>> Madhusudhan >> >
