Predicting Essential Genes in Fungal Genomes

Predicting Essential Genes in Fungal Genomes
By Michael Seringhaus†, Alberto Paccanaro†, Anthony Borneman, Michael Snyder and Mark Gerstein*

* To whom correspondence should be addressed: mark.gerstein@yale.edu
† These authors contributed equally to this work.

Supplementary Data Website "PredESS"
Main URL: http://www.gersteinlab.org/proj/predess/
Gerstein Group Homepage: http://www.gersteinlab.org/

Contents of this Website:

Supplementary Figures
Training the Machine Learning Classifier (S.cerevisiae)
Predicting Essential Genes (S.mikatae)
Supplementary Prediction Sets (S.pombe, S.bayanus)
Control: Randomization of 'ESSENTIAL' Label

1. Supplementary Figures

Naive Bayes - Orange Schema (image:PNG) -- BAYES-SCHEMA.png
Full table of essentiality predictions for S.mikatae (table:XLS) -- Smikatae_predictions.xls

2. Training the Machine Learning Classifier (S.cerevisiae)

S.cerevisiae

RAW FEATURES
- Scerevisiae/Features/
COMPILED FEATURES
- Scerevisiae/Compiled/
KNOWN ESSENTIAL GENES
- Scerevisiae/ESSENTIALGENES/

3. Predicting Essential Genes (S.mikatae)

S.mikatae

RAW FEATURES
- Smikatae/Features/
COMPILED FEATURES
- Smikatae/Compiled/
PREDICTED ESSENTIAL GENES
- Smikatae/Predictions/

4. Supplementary Prediction Sets (S.pombe, S.bayanus)

S.pombe

RAW FEATURES
- Spombe/Features/
COMPILED FEATURES
- Spombe/Compiled/
PREDICTED ESSENTIAL GENES
- Spombe/Predictions/

S.bayanus

RAW FEATURES
- Sbayanus/Features/
COMPILED FEATURES
- Sbayanus/Compiled/
PREDICTED ESSENTIAL GENES
- Sbayanus/Predictions/

5. Control: Randomization of 'ESSENTIAL' Label

As a control experiment, the 'essential' label was randomized among S.cerevisiae genes and our classifier was re-trained on this mock 'RANDOM-ESS' data.

MOCK TRAINING DATA for S.cerevisiae -- data/Scerevisiae/Mock-RandomESS-TrainingData-14-features/

This classifier was then applied to S.mikatae to predict 'essential' genes. These results are here:

MOCK PREDICTION DATA for S.mikatae -- data/Smikatae-vs-RANDOMESS/