| Main page |
Welcome to the "PredAlgo" server!
The website provides an interface to the PredAlgo program, a new multi-subcellular localization prediction tool dedicated to Algae. PredAlgo has been developed by Marianne Tardif, Ariane Atteia, Michael Specht, Cogne Guillaume, Norbert Rolland, Sabine Brugière, Michael Hippler, Myriam Ferro, Christophe Bruley, Gilles Peltier, Olivier Vallon, and Laurent Cournac, from various institutions mostly in France and in Germany (Tardif et al. 2012, Mol. Biol. Evol. 29: 3625-3639). The PredAlgo algorithm is based on a neural network trained to recognize transit peptide sequences. The neural network was trained using highly curated sets of proteins, for which subcellular locations were known, from the green alga Chlamydomonas reinhardtii. PredAlgo predicts whether a given protein is targeted to three pertinent compartments: the mitochondrion, the chloroplast, and the secretory pathway. It also predicts the length of transit peptide sequence based on the score profile, but this is not as reliable as the results of dedicated programs such as TargetP. Predictions are currently most reliable for Chlamydomonas and related green algae species (Chlorophyta).
Please, send any feedback and/or suggestions about this website to Olivier Vallon, Institut de Biologie Physico−Chimique (IBPC), Paris, France.
PredAlgo computations take a few seconds per sequence, and may take several minutes or hours on large datasets.
Note that many non-alphanumerical characters are not allowed in the sequence name, even if they are tolerated in the rest of the header: % : / $ # \ ^ ~ & é " ` @ µ £ ? ! = + *
If you receive a message "Invalid parameter in URL" or "Sequence(s) contain(s) weird characters", try removing them from your entry. For a large file, you can use the command (on Unix/Linux and Mac):
sed -e 's/@\// /g' INPUT.fasta | awk -F " " '{ if ($1 ~ /^>/) gsub("[\+\"\*\[\%\:\/\$\#\\\^\~\&\@\?é`µ£!\;]", "_", $1); print $0 }' > INPUT_corrected.fasta
You may also want to remove non alphabetical characters that may pollute your sequence by using:
awk '{ if ($1 !~ /^>/) gsub("[^A-Z]", "", $1); print $0 }'
If the issue remains, please contact us.