-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathrun_combine_downsample_pipeline.slurm
117 lines (86 loc) · 3.56 KB
/
run_combine_downsample_pipeline.slurm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
#!/bin/bash
#SBATCH --job-name=Downsample
#SBATCH --output=celltyping_combine_effect.out
#SBATCH --gres=gpu:1
source ~/.bashrc
PROJECT_PATH=""
ENV_PATH=""
conda activate $ENV_PATH/mars ## mars environment
cd $PROJECT_PATH
## fix method and feature selection
method="MLP"
FS="F-test"
##### =============== Mouse Brain Dataset
## set individual 2 as fixed target
test="P60FCCx3cr1Rep1"
declare -a train_list=("P60FCRep1" "P60FCRep2" "P60FCRep3" "P60FCRep4" "P60FCRep6" "P60FCAldh1l1Rep1")
## === Mousebrain FC each ind1 -> ind2
task="mousebrain_FC"
## === Mouse FC each ind1 -> ind2, with cell types curated (layers neuron -> neuron; two interneuron -> interneuron)
task="mousebrain_FC_curate"
#for train in ${train_list[@]}; do
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train $train --test $test
#done
## === Mousebrain FC ind1 -> ind2, using 81 sub-celltypes
task="mousebrain_FC_sub"
## Mousebrain FC exclude P60FCCx3cr1Rep1 -> P60FCCx3cr1Rep1
task="mousebrain_FC_multiinds"
train="ALL"
## Mousebrain FC exclude P60FCCx3cr1Rep1 -> P60FCCx3cr1Rep1, sub-cell types
task="mousebrain_FC_multiinds_sub"
train="ALL"
## Mousebrain FC exclude P60FCCx3cr1Rep1 -> P60FCCx3cr1Rep1, downsample to train cell numbers
task="mousebrain_FC_multiinds_sample"
train="Avg"
## Mousebrain FC exclude P60FCCx3cr1Rep1 -> P60FCCx3cr1Rep1, sub-cell types, downsample
task="mousebrain_FC_multiinds_sub_sample"
train="Avg"
#### ================ Mouse Brain across dataset
## TODO: wait on later
declare -a FC_train_list=("P60FCRep1" "P60FCRep2" "P60FCRep3" "P60FCRep4" "P60FCRep6" "P60FCAldh1l1Rep1" "P60FCCx3cr1Rep1") ## adult controls
declare -a pFC_test_list=("2" "3" "5" "7" "11" "12") ## saline samples, take as adult controls
## Mouse brain FC individual -> pFC Adult individual
task="mousebrain_FC_datasets_multiinds"
#for train in ${FC_train_list[@]}; do
# for test in ${pFC_test_list[@]}; do
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train "FC_${train}" --test "Adult_PFCSample${test}"
# done
#done
## Mouse brain all FC individual -> pFC Adult individuals
task="mousebrain_FC_datasets_multiinds"
train="FC"
#for test in ${pFC_test_list[@]}; do
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train $train --test "Adult_PFCSample${test}"
#done
task="mousebrain_FC_datasets_multiinds_sample"
train="FC"
#for i in {1..11}
#do
# RANDOM_NUM=$(($RANDOM % 1000))
# echo "Random downsampling seed, $RANDOM_NUM"
# for test in ${pFC_test_list[@]}; do
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train $train --test "Adult_PFCSample${test}" --sample_seed $RANDOM_NUM
# done
#done
##### =============== Human PBMC Dataset
## set individual S5 as fixed target
test="1085"
declare -a PBMC_train_list=("1043" "1079" "1154" "1249" "1493" "1511" "1598")
## Human PBMC batch1 ind1 -> ind2
task="PBMC_batch1_ind"
#for train in ${PBMC_train_list[@]}; do
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train $train --test $test
#done
## Human PBMC batch1 exclude ind2 -> ind2
task="PBMC_batch1_multiinds"
train="ALL"
## Human PBMC batch1 exclude ind2 -> ind2, downsample
task="PBMC_batch1_multiinds_sample"
train="ALL"
#for i in {1..30}
#do
# RANDOM_NUM=$(($RANDOM % 1000))
# echo "Random downsampling seed, $RANDOM_NUM"
# python -m pipelines.pipeline $task -m $method --select_on train --select_method $FS --train $train --test $test --sample_seed $RANDOM_NUM
#done
conda deactivate