Task 3 - Sequence-based predictions 2011
From Bioinformatikpedia
In contrast to the vast amount of known protein sequences, information about structure and function is available for only very few proteins. Sequence-based predictions of protein features aim to decrease this gap.
Contents
Introductory talks
Talk on sequence feature prediction: Media:Talk_Task3.pdf
The talk introduced the prediction of:
- secondary structure
- disorder
- transmembrane helices and signal peptides
- GO terms
Tasks and questions
For each prediction, carry out these assignments and answer these questions:
- Collect basic information about the methods: author, year, reference
- Give a brief description of the theory and the algorithm, if possible
- What is predicted? Describe the features in some detail
- What information is required for the predictions?
- Apply the prediction methods to your protein and explain how to do this
- You may try out other protein sequences (this is required when specified)
- Present, describe and discuss the results
- Look for other methods, for example here: http://expasy.org/tools/
- You may try out more methods
1. Secondary structure prediction
PSIPRED
PSIPRED (with hints on running psipred locally)
- /apps/psipred_2.5 and psipred_3.2
- Script: runpsipred -> adapt paths
- Readme: http://bioinfadmin.cs.ucl.ac.uk/downloads/psipred/
- Server: http://bioinf.cs.ucl.ac.uk/psipred/
Jpred3
Comparison with DSSP
- /apps/dssp
- Server: http://swift.cmbi.ru.nl/servers/html/ (-> Protein analysis -> Secondary Structure, symmetry and accessibility)
- http://swift.cmbi.ru.nl/gv/dssp
2. Prediction of disordered regions
DISOPRED
DISOPRED (with hints on running psipred locally)
- /apps/disopred_2.4.2.
- Script: rundisopred -> adapt paths
- Server: http://bioinf.cs.ucl.ac.uk/disopred/
POODLE
- Server: http://mbs.cbrc.jp/poodle/poodle.html
- for example, POODLE-I (POODLE series only)
IUPRED
- Server: http://iupred.enzim.hu/index.html
- Try different prediction types
META-Disorder
- Do not run locally. If you are interested, have a look at "man metadisorder" in the VM
- You can use the PredictProtein server for the prediction
3. Prediction of transmembrane alpha-helices and signal peptides
- Predict transmembrane helices and signal peptides additionally for the following proteins: BACR_HALSA, RET4_HUMAN, INSL5_HUMAN, LAMP1_HUMAN and A4_HUMAN
- You can look for other alpha-helical transmembrane proteins and proteins with different signal peptides to apply the prediction methods
- Why is the prediction of transmembrane helices and signal peptides grouped together here?
- Describe different signal peptides
TMHMM
- in /apps/tmhmm-2.0c/bin: adapt paths in tmhmm, tmhmm.ORIG and tmhmmformat.pl (/usr/local/bin/ -> /usr/bin/)
Phobius and PolyPhobius
- Server: http://phobius.sbc.su.se/
OCTOPUS and SPOCTOPUS
SignalP
- in /apps/signalp-3.0/signalp: adapt path
- run as superuser
TargetP
4. Prediction of GO terms
- Use also the additional proteins from the transmembrane/signal peptide prediction
GOPET
Pfam
- Database: http://pfam.sanger.ac.uk/ -> SEQUENCE SEARCH
ProtFun 2.2
TEST
line1 line2 line3 line4