Informatics

This page lists some of the informatics behind the work. The code presented here is open source, which allows scientific accountability of the Robot Scientist.

Statistics and data analysis

• Procedure and code for calculation of growth curve parameters. Results are listed in the Datalog data.
• Procedure and code to calculate statistics of growth curve parameters. Stats results can be found here.
• Procedure and code for discrimination (random forests) of growth curve parameters, and calculation of the significance of the discrimination. Results can be found here.
• Procedure and code for MC sampling for decision made when integrating results from trials. Results can be found here.

Data used to derive hypotheses

For BLAST searches we used:

• nr.00.tar.gz
• nr.01.tar.gz
• swissprot.tar.gz

from ftp://ftp.ncbi.nlm.nih.gov/blast/db/ (we downloaded these on 2nd Oct 2006) and the 2.2.15 version of blast from http://www.ncbi.nlm.nih.gov/BLAST/download.shtml (NOT netblast or wwwblast).

The KEGG data that was used to search for ORFs in other organisms.

Procedures for deriving hypotheses/experiments

• Procedure to automatically infer hypotheses relating enzymes and genes for novel biology and code to implement this.
• Procedure to automatically infer hypotheses relating enzymes and genes for the AAA pathway
• Procedure to automatically derive hypotheses (choose experiments) relating metabolites and genes from hypotheses relating enzymes and genes
• Procedure to layout experiment plates
• Procedure to automatically eliminate hypotheses based on integrated trial results for novel biology
• Procedure to automatically eliminate hypotheses based on integrated trial results for AAA pathway

Other code

• laboratory_control.tar.gz Code that designs experiments and lays out plates
• yeast_model_engine.tar.gz Model engine, for execution of model
• discovery_loops.tar.gz Code to execute the discovery loops for AAA pathway

Contact Ken Whelan at Aberystwyth for further information on this code.