Format 2 Description

This is a less common format used by one or two advanced servers and some simple hydrophobicity algorithm scripts. If your algorithm uses this type of result format, you will need to parse the output to match this format.

Each protein prediction starts with a greater than symbol > followed by the number corresponding to the ID given in tmheval.fa ranging from 0 to 2246. A prediction is assigned to each residue as a letter. Your algorithm may assign two-state predictions - H for helix, N for non-helix - or three state predictions including these two states and an intermediate or "questionable" state that is often found at the boundary between helical and non-helical residues.

2 state: NNNNNNNHHHHHHHHHHHHHHNNNNNNNNNNNNNNNNNNNNNN
3 state: NNNNNIIIHHHHHHHHHHHIINNNNNNNNNNIINNNNNNNNNN

You do not need to follow this letter code as you will be given the option to assign every letter used in your prediction file to be treated as a helical or non-helical residue after your file is uploaded. You will thus be able to change any 3-state predictions into 2-state predictions by assigning the intermediate state to be either helical or non-helical. Sample file format:

>0
LHHHHLLL
>1
LLLLLLLLLLHHHHH

Extra whitespace is generally not problematic:

OKAY:
> 0

LHHHHLLL

>1

LLLLLLLLLLHHHHH

But the record separator ">" must be at the beginning of a line

NOT OKAY:
 >0
LHHHHLLL
 >1
LLLLLLLLLLHHHHH

NOT OKAY:
>0
LHHHHLLL >1
LLLLLLLLLLHHHHH