fragment generation – local vs server

[estrauch]

Dear all

Can anyone clarify the advantages and disadvantages of these two ways of generating fragment libraries? I note that the FAQ now strongly recommends the server. Two advantages of the server might be

– access to secondary structure prediction methods not otherwise distributed (only Psipred seems available in a form that can be plugged into rosetta_fragments/nnmake/make_fragments.pl)

– automatic updating of nr etc

In general, though, it would be handy to be independent of server load and obtain fragments locally. My general experience is that make_fragments.pl does a good job.

Am I missing something? What method to other people prefer and why?

Cheers

Dan

[Anonymous]

Hi Dan,

We recommend the server because some users does not familiar with modifying the paths specified in perl scripts and the installations of other programs on linux, such as PSI_BLAST.

Generate fragments locally is more convenient if you know Perl scripts and familier with shell configurations.

Yi
> Dear all
>
> Can anyone clarify the advantages and disadvantages of these two ways of generating fragment libraries? I note that the FAQ now strongly recommends the server. Two advantages of the server might be
>
> – access to secondary structure prediction methods not otherwise distributed (only Psipred seems available in a form that can be plugged into rosetta_fragments/nnmake/make_fragments.pl)
>
> – automatic updating of nr etc
>
> In general, though, it would be handy to be independent of server load and obtain fragments locally. My general experience is that make_fragments.pl does a good job.
>
> Am I missing something? What method to other people prefer and why?
>
> Cheers
>
> Dan

[Author]

Posts