Looking for Python, Data Mining Expert -- 3

Tasks: 1. mine association rules using apriori ( min support - 0.05 , confidence - 20%) 2. Generalize each GO annotation in your input file by replacing with its parent as noted in the attached file ([login to view URL]). Mine rules after each [login to view URL] 3. the GO terms are hierarchial – mine relations. [login to view URL] tsv file – – replace it with direct parents – if repeats pick randomly – u have new set now – mine assoc rules . 5. 2nd round of replacement – opposite of 1st round of replacement – replace the already replaced parents with 2nd direct parents. Now you have 3 transaction sets. From each of 3 sets, mine assoc rules and put these rules together. 6. If u get rules x implies y, x implies (y ,z), just put x implies y. 4. If x implies y, if x is parent or y is parent, remove the rule- find these in text file. 5. If x is child of z, y implies z , infer the relation from go file.- get all the ancestors – say y and z are ancestors of x . 6. Filter any rules between an ancestor and child. Get ancestor children relationships from the [login to view URL] file. – remove x im y , y imp z , remove these both rules. Remaining – go terms that r not related .