run.pl

:- use_module(library(lists)).
:-dynamic state_of_entity_seen_before/3.
%:- dynamic tuple/3.
:- dynamic sorted_diff/1.
:-dynamic links/1.






%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%Class Predicates%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%This commented to surpess annoying warnings for the time being.

index_of_col(Col,Index):-
	collumns(List),
	List=[collumns|CoLList],
	indexOf(CoLList,Col,Index).



sample_diagnosis(Sample,Diagnosis,Tumor):-
	index_of_col('Diagnosis',Diagnosis_Index),
	index_of_col('Sample ID',Sample_Id_Index),
	index_of_col('Tumor stage',Tumor_stage_Index),
	sample_class(Sample_List),
	Sample_List=[sample_class|Sample_Tuple],
	Sample_Tuple=[H|_T],H\='Dataset ID',
	nth0(Diagnosis_Index,Sample_Tuple,Diagnosis),
	nth0(Sample_Id_Index,Sample_Tuple,Sample),
	nth0(Tumor_stage_Index,Sample_Tuple,Tumor).




%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%State related predicates%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%If a complex does not have any proteins, sets or complexes its state is on. Similar for inputs to a reaction. To be done.
%might find this too many times.. because of many -to many
probe_value_in_sample(Probe,Sample,Value):-
	samples(Sample, Probe, Value).

sample_tuple(Sample,Diagnosis,Tumor,Probe,Value):-
	samples(Sample, Probe,Value),
	sample_diagnosis(Sample,Diagnosis,Tumor).

%used where sample is instantiated. Finds a probe and value.
uniProt_Sample_Probes(Uniprot,Sample,Probe,Value):-
	samples2(Sample, Probe, Value, Uniprot).
	%old: sample_tuple(Sample,_D,_T,P,V),probe_uniProt(P,Uniprot). %I have switched this around.
		% probe_uniProt(P,'Q96T61'),sample_tuple('GSM102447',D,T,P,V).


sample(Sample):-
	%( nonvar(Dataset) ; Dataset = 'GSE2109' )
	Dataset = 'GSE2109', %default to 2109
	sample_class(List),
	List=[sample_class|Sample_Tuple],
	Sample_Tuple=[H|_T],H\='Dataset ID',
	index_of_col('Dataset ID', Dataset_Index),
	index_of_col('Sample ID',Sample_Index),
	nth0(Dataset_Index,Sample_Tuple,Dataset),
	nth0(Sample_Index,Sample_Tuple,Sample).


samples(Samples):-
	findall(Sample, sample(Sample),Samples). %defaulting to gse2109


samples_old(Samples):-
	setof(Sample_in, Probe^ProbeValue^Uniprot^samples2(Sample_in, Probe, ProbeValue, Uniprot), Samples).
    %member(Sample, Samples).
sample_old(Sample):-
	samples_old(Samples),
	%list_to_set(Samples, SetSamples),
	member(Sample, Samples).


protein_stack(Sample, Reactome_Id, State):-
        (
            protein_reactome_id_to_uniprot_id(Reactome_Id, Uniprot_Id),

            uniProt_Sample_Probes(Uniprot_Id, Sample, _Probe, 1),
            !,
            State is 1
        ;
            State is 0
        ).
%this works
protein(Sample, Reactome_Id,State):-
	sample(Sample),
	protein_reactome_id_to_uniprot_id(Reactome_Id, _),
	protein_stack(Sample, Reactome_Id,State).
	%write([Sample,Reactome_Id,State]).

%finds multiple copies but is fast
protein_new(Sample,Reactome_Id,State):-
	protein_reactome_id_to_uniprot_id(Reactome_Id,Uniprot_Id),
	samples2(Sample,_Probe,State,Uniprot_Id).



state_of_entity_seen_before(default,default,default).
retract_my_things:-
	retractall(state_of_entity_seen_before(_X,_Y,_Z)),
	asserta(state_of_entity_seen_before(default,default,default)). %put the default back in.

state_of_entity_tabled(Entity,State,Sample):-
	%try and find it in asserted if it is not in asserted, calculate it an assert it. Otherwise use it.
	(
	   not(state_of_entity_seen_before(Entity,State,Sample))
	->
	%if not there then calc it, and assert it.
	(
	  state_of_entity(Entity,State,Sample),
	  asserta(state_of_entity_seen_before(Entity,State,Sample))
	);
	  state_of_entity_seen_before(Entity,State,Sample)
	).


%untested
state_of_entity(Entity,State,Sample):-
	type(Entity, protein),
	protein_new(Sample, Entity,State),!.


%TODO need to make these faster
%small molecules are always 'on'
state_of_entity(Entity,State,_Sample):-
	component_type(Entity, 'SmallMolecule'),
	State is 1,!.
%Physical entities are always 'on'
state_of_entity(Entity,State,_Sample):-
	component_type(Entity, 'PhysicalEntity'),
	State is 1,!.




%if a (simple) entity is a protein set and its state is on
state_of_entity(Entity,State,Sample):-
	type(Entity, protein_set),
	all_children_proteins(Entity),
	child_component(Entity,Child),
	protein_new(Sample,Child,1),
	State is 1,!.

%if a (simple) entity is a protein set and it's state if off
%I find all proteins for a sample, this is a list of values, I delete all the
%zeros and the remaining list will unify with the empty list.
state_of_entity(Entity,State,Sample):-
	type(Entity, protein_set),
	all_children_proteins(Entity),
	child_component(Entity,Child),
	bagof(Value, Value^protein_new(Sample,Child,Value),Vs),
	delete(Vs,0,ListOfOnes),ListOfOnes=[],
	State is 0,!.

%if a (simple) entity is a complex and is off
state_of_entity(Entity,State,Sample):-
	type(Entity, complex),
	all_children_proteins(Entity),
	child_component(Entity,Child),
	protein_new(Sample,Child,0),
	State is 0,!.
%if a (simple) entity is a complex and is on.
%I find all protein in a sample, this is a list of values, I delete all the
%zeros and the remaining list will unify with the empty list.
state_of_entity(Entity,State,Sample):-
	type(Entity, complex),
	all_children_proteins(Entity),
	child_component(Entity,Child),
	bagof(Value, Value^protein_new(Sample,Child,Value),Vs),
	delete(Vs,1,ListOfZeros),ListOfZeros=[],
	State is 1,!.

%if a complex with components is off
state_of_entity(Entity,State,Sample):-
	type(Entity, complex),
	child_component(Entity,Child),
	(state_of_entity(Child,0,Sample);
	protein_new(Sample,Child,0)), %if it has any proteins as input as well as other components
	State is 0,!.

%if a complex with components is on
state_of_entity(Entity,State,Sample):-
	type(Entity, complex),
	child_component(Entity,Child),
	bagof(Value, Value^state_of_entity_tabled(Child,Value,Sample),Vs),%if it has component inputs
	bagof(Value2, Value2^protein_new(Sample,Child,Value2),Vs2),%if it has protein inputs
	append(Vs, Vs2, Vs3),
	delete(Vs3,1,ListOfZeros),ListOfZeros=[],%delete all the ones, the list of zeros will be empty if all inputs are on
	State is 1,!.

%if a protein set with components is on
state_of_entity(Entity,State,Sample):-
	type(Entity, protein_set),
	child_component(Entity,Child),
	(state_of_entity_tabled(Child,1,Sample);
	protein_new(Sample,Child,1)), %if it has any proteins as input as well as other entities
	State is 1,!.

%if a protein set with components is off
state_of_entity(Entity,State,Sample):-
	type(Entity, protein_set),
	child_component(Entity,Child),
	bagof(Value, Value^state_of_entity_tabled(Child,Value,Sample),Vs), %if it has entity inputs
	bagof(Value2, Value2^protein_new(Sample,Child,Value2),Vs2), %if it has protein inputs
	append(Vs, Vs2, Vs3), %join the list of inputs together
	delete(Vs3,0,ListOfOnes),ListOfOnes=[], %delete all the zeros, the list of 1's will be empty if all inputs are off
	State is 0,!.



%We will need two more for state of inputs to a reaction %all inputs need to be present. if controlled by activation- needs to be activated if controlled by inhibiton can not be inhibited
%reaction_can_procceed(Reaction, State,Sample):-
reaction_inputs_states(Reaction, Sample, Input_States):-
	inputs_to_reaction(Reaction,Inputs_List),
	findall(State,(member(X,Inputs_List), state_of_entity_tabled(X,State,Sample)), Input_States).

reaction_all_inputs_on(Reaction,Sample,State):-
	reaction_inputs_states(Reaction, Sample, Input_States),
	delete(Input_States,1,ListOfZeros),ListOfZeros=[], %We delete all the ones from the input states, to give us a list of zeros, if this is an empty list then all inputs were on
	State is 1,!.

reaction_all_inputs_on(Reaction,Sample,State):-
	reaction_inputs_states(Reaction, Sample, Input_States),
	delete(Input_States,1,ListOfStates),
	member(_X,ListOfStates), %if there is a input that was not on
	State is 0,!.

%%	TODO NEED TO CHECK FOR FUNNY ACTIVIAITON THINGS

%if a reaction is controlled by an activation this must be on
%!!!Reaction must be instansinated for this to work correctly!!!%
%reaction_activated(+Reaction,Sample)
reaction_activated(Reaction,Sample):-
	reaction_controlled_by(Reaction,_ControlId,Controller_Id, 'ACTIVATION'),
	state_of_entity_tabled(Controller_Id,1,Sample).
	%writeln('Activated').	%is this properly tested?

%if a reaction is not controlled by an activation then we say it is active.
%!!!Reaction must be instansinated for this to work correctly!!!%
%reaction_activated(+Reaction,Sample)
%this is quite slow
reaction_activated(Reaction,_Sample):-
	not(reaction_controlled_by(Reaction,_ControlId,_Controller_Id, 'ACTIVATION')).
	%writeln('Activated'). %reaction is not controlled by an activation so defaults to being activated

%if a reaction is controlled by an inhibition this must be off (does not work with not)
reaction_inhibited(Reaction,Sample):-
	reaction_controlled_by(Reaction,_ControlId,Controller_Id, 'INHIBITION'),
	state_of_entity_tabled(Controller_Id,1,Sample).
	%writeln('Inhibited').

%reaction_can_procceed
reaction_can_procceed(Reaction,Sample):-
	reaction_activated(Reaction,Sample),
	not(reaction_inhibited(Reaction,Sample)),
	reaction_all_inputs_on(Reaction,Sample,1).


reaction_can_procceed(Reaction,Sample,State):-
		reaction_can_procceed(Reaction,Sample),
		State is 1,!.

reaction_can_procceed(_Reaction,_Sample,State):-
	   State is 0,!.

%11:06

mini_sample(Sample):-
	member(Sample, ['GSM102447','GSM102449','GSM102451','GSM102455','GSM102507']).


%%%%%%Timeing preds
%
time_of_reactions(Time1,Time2, _Sample, R, State):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	%rdf(Rs,'http://www.w3.org/1999/02/22-rdf-syntax-ns#type','http://www.biopax.org/release/biopax-level3.owl#BiochemicalReaction'), %checking R is biochemical reaction
	%reactome_string_id(Rs, R),
	reaction_can_procceed(R,'GSM102449',State),
	statistics(runtime,[_,Time2]).




make_a_run(File,Pathway,Set,Time1,Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),
	forall(
	    (
	     sample(Sample),
	     Sample\='GSM89060', %problem with this as not in csv file
	     reaction_in_pathway(Pathway, Set,Reaction), %this is just a generator at the moment
	     reaction_can_procceed(Reaction,Sample,State)
	    ),
	    (
	     format("tuple('~w','~w',~w). ~n",[Sample,Reaction,State]),
	     format(Stream,"tuple('~w','~w',~w). ~n",[Sample,Reaction,State])
	    )
	),

	close(Stream),
	statistics(runtime,[_,Time2]).

all_samples_all_reactions(File,Time1,Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),
	forall(
	    (	sample(Sample),Sample\='GSM89060'), %problem with this as not in csv file
	    (
		forall(
		    (
		    component_type_slow(Reaction,'BiochemicalReaction'),

		    reaction_can_procceed(Reaction,Sample,State)
		),
		    (
			format("tuple('~w','~w',~w). ~n",[Sample,Reaction,State]),
			format(Stream,"tuple('~w','~w',~w). ~n",[Sample,Reaction,State])
		    )
		)
	    )
	),
	close(Stream),
	statistics(runtime,[_,Time2]).

all_reactions_in_sample(Stream,Sample):-
	Sample\='GSM89060',
	forall(
	    (
	      component_type_slow(Reaction,'BiochemicalReaction'),
	      reaction_can_procceed(Reaction,Sample,State)
	     ),
	    (
		format("tuple('~w','~w',~w).~n",[Sample,Reaction,State]),
		format(Stream,"tuple('~w','~w',~w).~n",[Sample,Reaction,State])
	    )
	).



myThread(Stream,Sample):-all_reactions_in_sample(Stream,Sample),thread_exit(Sample).


start3(File,Time1,Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),
	findall(myThread(Stream,Sample),(sample(Sample),Sample\='GSM89060'),Goals),
	concurrent(8,Goals,[]),
	close(Stream),
	statistics(runtime,[_,Time2]).


start_batch(File,Batch,Time1,Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),
	findall(myThread(Stream,Sample),(batch(Batch,Sample)),Goals),
	concurrent(8,Goals,[]),
	close(Stream),
	statistics(runtime,[_,Time2]).


start_batch_non_concurrent(File,Batch,Time1,Time1):-
	open(File,append,Stream,[]),
	findall(Sample,batch(Batch,Sample),Samples),
	forall(
	    member(S,Samples),
	    all_reactions_in_sample(Stream,S)
	),
	close(Stream).





one_sample_all_reactions(File,Sample,Time1,Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),

	forall(
	    (
	     component_type_slow(Reaction,'BiochemicalReaction'),
	     Sample\='GSM89060', %problem with this as not in csv file
	     reaction_can_procceed(Reaction,Sample,State)
	    ),
	    (
	     format("tuple('~w','~w',~w). ~n",[Sample,Reaction,State]),
	     format(Stream,"tuple('~w','~w',~w). ~n",[Sample,Reaction,State])
	    )
	),


	close(Stream),
	statistics(runtime,[_,Time2]).


stats(File,Sample,Time1, Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),

	forall(
	    (component_type_slow(Pathway_Id,'Pathway')
	    ),
	    (
		reactome_name(Pathway_Id, Pathway_Name),
		findall(State,(reaction_in_pathway(Pathway_Id, _Set,Reaction),tuple(Sample,Reaction,State)),States),
		length(States,Length),
		sum_list(States,Sum),
		format("~w#~w#~w#~w ~n",[Pathway_Id,Pathway_Name,Length,Sum]),
		format(Stream,"~w#~w#~w#~w ~n",[Pathway_Id,Pathway_Name,Length,Sum])
	    )
	),
	close(Stream),
	statistics(runtime,[_,Time2]).


stats_compare(File,Sample1,Sample2,Time1, Time2):-
	statistics(runtime,_),
	statistics(runtime,[_,Time1]),
	open(File,append,Stream,[]),

	forall(
	    (component_type_slow(Pathway_Id,'Pathway')
	    ),
	    (
		reactome_name(Pathway_Id, Pathway_Name),
		findall(State,(reaction_in_pathway(Pathway_Id, _Set,Reaction),tuple(Sample1,Reaction,State)),States),
		findall(State2,(reaction_in_pathway(Pathway_Id, _Set2,Reaction2),tuple(Sample2,Reaction2,State2)),States2),
		length(States,Length),
		%length(States2,Length2),
		%Length2 \=0,
		%Length \=0,
		%Length >=10,
		sum_list(States,Sum),
		%Percentage1 is Sum/Length,
		sum_list(States2,Sum2),
		%Percentage2 is Sum2/Length,
		%Value is abs(Percentage1-Percentage2),
		%Value >=0.3,

		format("~w#~w#~w#~w#~w ~n",[Pathway_Id,Pathway_Name,Length,Sum,Sum2]),
		format(Stream,"~w#~w#~w#~w#~w ~n",[Pathway_Id,Pathway_Name,Length,Sum,Sum2])
	    )
	),
	close(Stream),
	statistics(runtime,[_,Time2]).






%%%%%%CSV writing
%
%reactions in tuple set
%WARNING CHECK TUPLE IS CORRECT FIRST. retractall(tuple(_,_,_)).
%
%
%
pathway_enriched(Sample,Pathway_Id,Total,Total_On,Percent_On):-
	batch(_,Sample),
	component_type_slow(Pathway_Id,'Pathway'),
	all_reactions_in_pathway(Pathway_Id,Reactions),
	length(Reactions,Total),
	Total \=0, %some pathways have no reactions? !
	findall(State,(reaction_in_pathway(Pathway_Id, _Set,Reaction),tuple(Sample,Reaction,State)),States),
	sum_list(States,Total_On),
	Percent_On is Total_On/Total.

%component_type_slow(Pathway_Id,'Pathway').

pathway_enriched_avg(Class,Pathway_Id,Avg_On):-
	findall(Sample,(batch(_,Sample),sample_diagnosis(Sample,Class,_)),Samples),
	findall(Percent_On,(member(Single,Samples),pathway_enriched(Single,Pathway_Id,_,_,Percent_On)),Percent_On_List),
	sum_list(Percent_On_List,Sum_Percent_On),
	length(Percent_On_List,Number_Of_Reactions),
	Number_Of_Reactions \=0,
	Avg_On is Sum_Percent_On/Number_Of_Reactions.


pathway_enriched_diff_c(Pathway_Id,Diff):-%calculated
	component_type_slow(Pathway_Id,'Pathway'),
	pathway_enriched_avg('AC',Pathway_Id,AC_Avg_On),
	pathway_enriched_avg('SCC',Pathway_Id,SCC_Avg_On),
	Diff is abs(AC_Avg_On - SCC_Avg_On).

all_pathway_diffs(File):-
	open(File, append,Stream,[]),
	   forall(
	         pathway_enriched_diff(Pathway_Id,Diff),
	         (
		     format("pathway_enriched_diff('~w',~w).~n",[Pathway_Id,Diff]),
	             format(Stream,"pathway_enriched_diff('~w',~w).~n",[Pathway_Id,Diff])
		 )

	   ),

	close(Stream).

%%%%%%%%%%%%%%%Custom Sort%%%%%%%%%%%%%%%
%gt(X,Y):-X>Y.

gt(pathway_enriched_diff(_,X),pathway_enriched_diff(_,Y)):-X>Y.

bubblesort_p(List,Sorted):-
	swap(List,List1),!,
	bubblesort_p(List1,Sorted).

bubblesort_p(Sorted,Sorted).

swap([X,Y|Rest],[Y,X|Rest]):-
	gt(X,Y).

swap([Z|Rest],[Z|Rest1]):-
	swap(Rest,Rest1).

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

sort_pathway_diff(Sorted):-
	findall(pathway_enriched_diff(Pathway_Id,Diff),pathway_enriched_diff(Pathway_Id,Diff),AllDiff),
	bubblesort_p(AllDiff,Sorted),
	retractall(sorted_diff(_)),
	reverse(Sorted,SortedR),
	assert(sorted_diff(SortedR)).

pathway_with_high_diff(Min_Size,Pathway_Id,Diff):-
	sorted_diff(Sorted),
	member(pathway_enriched_diff(Pathway_Id,Diff),Sorted),
	all_reactions_in_pathway(Pathway_Id, Reactions),
	length(Reactions,L),
	L> Min_Size.


type_for_aleph(File):-
	open(File,append,Stream,[]),
	    forall(
		batch(_,Sample),
		format(Stream,"sample('~w').~n",[Sample])
	    ),
	    format(Stream,"~n~n~n",[]),
	    forall(
		reaction(R),
		format(Stream,"reaction('~w').~n",[R])
	    ),

	close(Stream).

aleph_models(File,Pathway):-
	open(File,append,Stream,[]),
	forall(
	    batch(_X,Sample),
	    (
		format(Stream,"%% microarray('~w').~n~n",[Sample]),
	        forall(
		    (
		    all_reactions_in_pathway(Pathway, Reactions),
		    member(Reaction,Reactions)
		     ),
		    (
			tuple(Sample,Reaction,State),
		        format(Stream,"reaction('~w','~w',~w).~n",[Sample, Reaction,State])
		    )),

	        format(Stream,"~n~n",[])
	    )
	),
	close(Stream).



warmr_models(File,Pathway):-
	open(File,append,Stream,[]),
	forall(
	    batch(_X,Sample),
	    (
		format(Stream,"begin(model('~w')).~n~nmicroarray('~w').~n~n",[Sample,Sample]),
	        forall(
		    (
		    all_reactions_in_pathway(Pathway, Reactions),
		    member(Reaction,Reactions)
		     ),
		    (
			tuple(Sample,Reaction,State),
		        format(Stream,"reaction('~w','~w',~w).~n",[Sample, Reaction,State])
		    )),
		sample_diagnosis(Sample,Class,_),
		downcase_atom(Class,Low_Class),
		format(Stream,"~w.~n~n",[Low_Class]),

	        format(Stream,"end(model('~w')).~n~n",[Sample])
	    )
	),
	close(Stream).

warmr_models_all(File):-
	open(File,append,Stream,[]),

	forall(
	    batch(_X,Sample),
	    (
		format(Stream,"begin(model('~w')).~n~nmicroarray('~w').~n~n",[Sample,Sample]),
	        forall(
		    (

		    reaction(Reaction)
		     ),
		    (
			tuple(Sample,Reaction,State),
		        format(Stream,"reaction('~w','~w',~w).~n",[Sample, Reaction,State])
		    )),
		sample_diagnosis(Sample,Class,_),
		downcase_atom(Class,Low_Class),
		format(Stream,"~w.~n~n",[Low_Class]),

	        format(Stream,"end(model('~w')).~n~n",[Sample])
	    )
	),
	close(Stream).





reaction(R):-
	setof(R,Sample^State^tuple(Sample,R,State),Rs),
	member(R,Rs).


write_csv_to_screen:-
	write('samples'),
	foreach(reaction(R),(write(','),write(R))),
	write(',class'),
	writeln(''),
	forall((sample(Sample),
	       Sample\='GSM89060'), %problem with this as not in csv file

		(   write(Sample),
		    foreach(tuple(Sample,R,S),(write(','),write(S))),
		    %class(Sample,Class),
		    sample_diagnosis(Sample,Class,_),
		    write(','),write(Class),writeln('')
		)).

write_csv_to_file(File):-
	open(File,append,Stream,[]),
	write(Stream,'samples'),
	foreach(reaction(R),(write(Stream,','),write(Stream,R))),
	write(Stream,',class'),
	nl(Stream),
	forall((sample(Sample),
	       Sample\='GSM89060'), %problem with this as not in csv file

		(   write(Stream,Sample),
		    foreach(tuple(Sample,R,S),(write(Stream,','),write(Stream,S))),
		    %class(Sample,Class),
		    sample_diagnosis(Sample,Class,_),
		    write(Stream,','),write(Stream,Class),nl(Stream)
		)),
	close(Stream).


write_aleph_pos(File):-
	open(File,append,Stream,[]),
	    forall(
	       (
		 batch(_X,Sample),
	         sample_diagnosis(Sample,'AC',_)
	       ),

	      (
		format(Stream,"microarray('~w').~n",[Sample])
	      )
	    ),


	close(Stream).



write_aleph_neg(File):-
	open(File,append,Stream,[]),
	    foreach(
		(
		batch(_X,Sample),
		sample_diagnosis(Sample,'SCC',_)
	       ),

	      (
				format(Stream,"microarray('~w').~n",[Sample])
	      )
	    ),


	close(Stream).



batch(1,'GSM102447').
batch(1,'GSM102449').
batch(1,'GSM102451').
batch(1,'GSM102455').
batch(1,'GSM102507').
batch(1,'GSM102512').
batch(1,'GSM102548').
batch(1,'GSM102553').
batch(2,'GSM102555').
batch(2,'GSM117610').
batch(2,'GSM117629').
batch(2,'GSM117632').
batch(2,'GSM117763').
batch(2,'GSM117770').
batch(2,'GSM137910').
batch(2,'GSM137912').
batch(3,'GSM137916').
batch(3,'GSM137931').
batch(3,'GSM137945').
batch(3,'GSM138002').
batch(3,'GSM138003').
batch(3,'GSM152594').
batch(3,'GSM152609').
batch(3,'GSM152624').
batch(4,'GSM152670').
batch(4,'GSM152681').
batch(4,'GSM152757').
batch(4,'GSM179786').
batch(4,'GSM179827').
batch(4,'GSM179918').
batch(4,'GSM179951').
batch(4,'GSM203641').
batch(5,'GSM203643').
batch(5,'GSM203699').
batch(5,'GSM203732').
batch(5,'GSM231874').
batch(5,'GSM231885').
batch(5,'GSM231897').
batch(5,'GSM231899').
batch(5,'GSM231907').
batch(6,'GSM231950').
batch(6,'GSM231952').
batch(6,'GSM277678').
batch(6,'GSM277695').
batch(6,'GSM277696').
batch(6,'GSM301680').
batch(6,'GSM353885').
batch(6,'GSM353933').
batch(7,'GSM38100').
batch(7,'GSM38103').
batch(7,'GSM38104').
batch(7,'GSM46817').
batch(7,'GSM46824').
batch(7,'GSM46833').
batch(7,'GSM46850').
batch(7,'GSM46868').
batch(8,'GSM46884').
batch(8,'GSM46936').
batch(8,'GSM46941').
batch(8,'GSM46973').
batch(8,'GSM46976').
batch(8,'GSM53167').
batch(8,'GSM53170').
batch(8,'GSM76585').
batch(9,'GSM76587').
batch(9,'GSM76590').
batch(9,'GSM76595').
batch(9,'GSM88962').
batch(9,'GSM88997').
batch(9,'GSM89046').