In order to judge the quality of a data model for Drools you should have a fairly conclusive idea about the rules you are going to write. (This isn't different from other applications: Storing the data in some organisation may work fine, but fitness for access is a different matter.)
The Lists may stand investigation. Is the order in the List important? Is the List just a hidden set? Is an m:n relationship better expressed by an entity of its own? The relational data model of DBs is a good approach for being prepared for a variety of requests and, therefore, rules. -W On 16/05/2014, lindenb <[email protected]> wrote: > Hi all, > > I'm working in the field of bioinformatics where I study more a large > number > of variations (N> 10E6) along the human genome among individuals affected > by a genetic disease. ( * ) > > my data looks like this (much simplified !) > > #CHROMOSOME POSITION GENE PROPERTIES SAMPLE1 SAMPLE2 > SAMPLE3 > chr1 987 GENE1,GENE3 score=1 null A/A A/G > chr1 988 GENE3 score=4;id=989 A/G null A/A > chr1 1988 null score=4;id=989 C/G null C/A > > > People in my lab have to filter those variation using different strategy, > they then extract the overlapping genes and then filter those genes to find > one or more gene that could explain the disease. > > Currently they use knime.org to run those filters. > (http://www.myexperiment.org/workflows/2320.html ) > > But I've got the feeling that drools could be used to filter those > variations. > Users would 'just' have to write some rules, run the engine and get the > result. > Rules would be removed/switched to get the result for another strategy. > > Question: > 1) is drools a good choice here ? > 2) how should I model my data ? > > If I use this java model: > > class Gene { > List<Variation> getVariations(); > } > > class Variation { > String getChromosome(); > int getPosition(); > List<Gene> getGenes(); > List<Genotype> getGenotypes(); > } > > class Genotype{ > Variation getVariation(); > Sample getSample(); > String getDNA(); > } > > it will create a large graph, can drools support it ? > > 3) Where in the documentation can I find how to easily set the state my > objects in order to filter them. E.g: > > variations ->[filters1]-> [filter2]----------------->[fiilter5] -> result > `>[filter3]->[filter4]--/ > > Thank you, > > Pierre > > (*) > http://www.1000genomes.org/wiki/Analysis/Variant%20Call%20Format/vcf-variant-call-format-version-41 > > > > -- > View this message in context: > http://drools.46999.n3.nabble.com/Newbie-is-my-data-model-suitable-for-drools-tp4029590.html > Sent from the Drools: User forum mailing list archive at Nabble.com. > _______________________________________________ > rules-users mailing list > [email protected] > https://lists.jboss.org/mailman/listinfo/rules-users > _______________________________________________ rules-users mailing list [email protected] https://lists.jboss.org/mailman/listinfo/rules-users
