norsnet(1) identifies unstructured loops from sequence




NORSnet is a neural network based method that focuses on the identification of unstructured loops.

NORSnet was trained to distinguish between very long contiguous segments with non-regular secondary structure (NORS regions) and well-folded proteins. NORSnet was trained on predicted information rather than on experimental data. Therefore, it was optimized on a large data, which is not biased by today's experimental means of capturing disorder. Thus, NORSnet reached into regions in sequence space that are not covered by the specialized disorder predictors. One disadvantage of this approach is that it is not optimal for the identification of the ``average'' disordered region.

Conversion of PSI-BLAST alignment to HSSP format

The most up-to-date procedure can be found at <>.
1. Convert BLAST output to a Single Alignment Format (SAF):
 /usr/share/librg-utils-perl/ fasta=<query_fasta_file> maxAli=3000 eSaf=1 \
  saf=<saf_formatted_file> <blast_output>
2. Convert SAF format to HSSP:
 /usr/share/librg-utils-perl/ <saf_formatted_file> formatIn=saf formatOut=hssp \
  fileOut=<hssp_formatted_file> exeConvertSeq=convert_seq
3. Filter results to 80% redundancy:
 /usr/share/librg-utils-perl/ red=80 <hssp_formatted_file> fileOut=<filtered_hssp_formatted_file>

Output format

Output mode 1

Tabular output, columns:

 pos            amino acid number (1..)
 res            residue 1-letter code
 node1          output of neural network node 1
 node2          output of neural network node 2
 pred           node1 / ( node1 + node2 )
 n40            pred < 0.40 ? '-' : 'N'
 n40fil         at least 31 AA long stretches of 'N' in n40
 n59            pred < 0.59 ? '-' : 'N'
 n59fil         at least 31 AA long stretches of 'N' in n59

'N' is for non-regular secondary structure.


Schlessinger, A., Liu, J., and Rost, B. (2007). Natively unstructured loops differ from other loops. PLoS Comput Biol, 3(7), e140.


File containing protein amino-acid sequence in fasta format.
Secondary structure and solvent accessibility prediction by PROF in rdb format.
PSI-BLAST alignment profile file converted to HSSP format.
The name of the final NORSnet output file.
Flexible/rigid residues prediction by profbval(1) in rdb format (mode 5).
NORSnet can create output files in different formats for different purposes. Valid modes are `1', `2' or `3'. Default mode: 1.
Default mode. Use this when you do not want to give a value here but you want to specify debug.
for metadisorder(1)
Set to 1 for debugging messages


number -
residue number
residue -
residue type
raw -
raw value of the different between the two output nodes


 norsnet /usr/share/doc/norsnet/examples/cad23.f /usr/share/doc/norsnet/examples/cad23-fil.rdbProf /usr/share/doc/norsnet/examples/cad23-fil.hssp cad23.norsnet cad23 /usr/share/doc/norsnet/examples/cad23.profbval


Overrides /usr/share/norsnet, the path to helper scripts and data files.


default output file extension
default precomputed input files directory


1. It is recommended to create the profiles using 3 iteration of PSI-BLAST against big database
2. It is also recommended to filter the hssp files using from the Prof package using the following command: perl hssp_file red=80


A. Schlessinger <[email protected]>