Re: [scipion-users] Local Resolution estimation(MonoRes and DeepRes) and Local Deblur

[JOSE L. V. P. <jl...@cn...>]

Dear Fan,

Thanks for getting in touch with us.

The input of localdeblur can be any local resolution map, it means  
that you can compute the local resolution with monores, resmap,  
blocres, o deepres. There is no one better than the other.

In the case of monores, as well as resmap, it is mandatory that your  
map contains noise, if it is noise free, use as I put  two halves. It  
should always be the best way of estimating the local resolution.

Also, deepres is a good choice, because it overcomes the noise  
problem, and you can computer the local resolution map of a noise-free  
density map.

Deepres belongs to xmipp package and you should be able to find it in  
analisys tab in the left panel. Alternatively, you can do crtl+F and  
search by writing the name of the protocol, it means deepres.

Regarding with the micelle, you can create a mask and apply it to remove it

Anything you need, please let us know

Kind regards

José Luis Vilas

Quoting 范宏成 <187...@16...>:

> Dear Colleagues,
> I have some questions about Local resolution estimation and Local  
> Deblur.  Looking forward to your response . Thank you!
> 1. MonoRes. Can MonoRes estimate the local resolution when I do  
> focused refinement (masked refinement) using relion. I mean there  
> maybe no region  to estimate the noise.  In this condition, I have  
> to use the relion local resolution protocol to get the focused map  
> local resolution.
> 2. Local Deblur.  Can Local Deblur support the relion local  
> resolution map or only support MonoRes,Resmap, Blocres?
> 3. DeepRes. I recently read a paper about the DeepRes and find it  
> very useful for my project.  I didn't find the DeepRes protocol in  
> my scipion. My scipion version is v2.0(2019-04-23) Diocletian in  
> Centos7 system(Binary version).  How can I use this protocol?
> 4.DeepRes for membrane protein. In DeepRes paper, It said it is very  
> important to mask out the micelle because DeepRes only trained with  
> proteins.  If I don't mask out the micelle, What bad consequencies  
> will happen? By the way, How to effectively and carefully  mask out  
> the micellce? In my way, I deleted the micelle using particle  
> substraction and run again auto-refinement in relion. In order not  
> to delete the signal from the protein, I used to use the soft mask,  
> So My membrante protein  had a small number of micelles around the  
> TM domain. I'm afraid this residual micelle may cause bad  
> consequencies when using DeepRes. Does someone have a better way to  
> delete the micelle in membrane proteins?
>
>
> Fan Hongcheng
>
>
> | |
> 范宏成
> |
> |
> m18...@16...
> |
> 签名由网易邮箱大师定制