Skip to content

Commit

Permalink
scenic+ temp_dir changed
Browse files Browse the repository at this point in the history
  • Loading branch information
@janursa
janursa committed Aug 28, 2024
1 parent d2a82e2 commit 3282935
Show file tree
Hide file tree
Showing 24 changed files with 661 additions and 276 deletions.
Binary file added hg38_screen_v10_clust.regions_vs_motifs.rankings.feather.aF53FD8D
View file
Binary file not shown.
Binary file added hg38_screen_v10_clust.regions_vs_motifs.scores.feather.e391A0Ff
View file
Binary file not shown.
283 changes: 90 additions & 193 deletions runs.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3432,9 +3432,7 @@
"metadata": {},
"outputs": [],
"source": [
"n_iter = 100\n",
"# net_corr = net.sample(len(tfs), axis=1)\n",
"# net_corr"
"n_iter = 100"
]
},
{
Expand Down Expand Up @@ -5226,252 +5224,151 @@
},
{
"cell_type": "code",
"execution_count": 57,
"execution_count": 59,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"[0.6338300913837829,\n",
" 0.6306478594813784,\n",
" 0.6326987253863688,\n",
" 0.6343378657887133,\n",
" 0.6326487820571753,\n",
" 0.6356536986072162,\n",
" 0.6301186542660742,\n",
" 0.6333196717157581,\n",
" 0.6273994641913987,\n",
" 0.6308608591441454,\n",
" 0.6289856651506277,\n",
" 0.6324828084784677,\n",
" 0.6334077063898018,\n",
" 0.6330415245102411,\n",
" 0.6333819774731821,\n",
" 0.6338439982092101,\n",
" 0.6319344177161561,\n",
" 0.6341557592562631,\n",
" 0.6307591889466913,\n",
" 0.6322247379885256,\n",
" 0.6306792388537548,\n",
" 0.6306696924928431,\n",
" 0.6285119151639946,\n",
" 0.6304831829144517,\n",
" 0.6308121613551088,\n",
" 0.6327618388683754,\n",
" 0.6314231674476822,\n",
" 0.6312884282584914,\n",
" 0.631142772964502,\n",
" 0.6320765893840486,\n",
" 0.6325844910761261,\n",
" 0.6301334652402374,\n",
" 0.6318591760956985,\n",
" 0.6320593747906565,\n",
" 0.6331810761679495,\n",
" 0.6314261332296257,\n",
" 0.6319687151358445,\n",
" 0.6307678470102952,\n",
" 0.6281131512485407,\n",
" 0.6348718405843499,\n",
" 0.6291553093895569,\n",
" 0.6303376046037746,\n",
" 0.6330107358097978,\n",
" 0.6351402003670711,\n",
" 0.6346943470719033,\n",
" 0.6286424673972038,\n",
" 0.6339761428059248,\n",
" 0.6276169663820544,\n",
" 0.6286235357673238,\n",
" 0.6342169461721089,\n",
" 0.6290226581971139,\n",
" 0.6314228512414537,\n",
" 0.6311366707045039,\n",
" 0.6327379499422254,\n",
" 0.6300788825982234,\n",
" 0.6334970039445497,\n",
" 0.6315234639260399,\n",
" 0.631137721674292,\n",
" 0.6322800527494545,\n",
" 0.6310507562501161,\n",
" 0.6327209453667971,\n",
" 0.627939229431186,\n",
" 0.6331771446984572,\n",
" 0.6335969281489019,\n",
" 0.6311342753756505,\n",
" 0.6318286221517255,\n",
" 0.6302232282749309,\n",
" 0.6289309088132349,\n",
" 0.6290424006853144,\n",
" 0.6337237311185296,\n",
" 0.6277177540589839,\n",
" 0.6291169600379366,\n",
" 0.6320369605246907,\n",
" 0.6280705657169147,\n",
" 0.6329689969853456,\n",
" 0.6285159998085228,\n",
" 0.6293034934160794,\n",
" 0.6327535340594785,\n",
" 0.6314765267336526,\n",
" 0.6355853836785672,\n",
" 0.6318860982148191,\n",
" 0.6326334172009915,\n",
" 0.6318694762265682,\n",
" 0.6335425559144152,\n",
" 0.6339342326320021,\n",
" 0.6297642298354758,\n",
" 0.6265141566138808,\n",
" 0.6340795309921302,\n",
" 0.6315806409233686,\n",
" 0.6328343170100053,\n",
" 0.6315533175464346,\n",
" 0.6314159987840019,\n",
" 0.6342163661185687,\n",
" 0.6323350759040771,\n",
" 0.6303224210995871,\n",
" 0.6312488527545446,\n",
" 0.6326553474980217,\n",
" 0.6336276777942278,\n",
" 0.6323158375908737]"
"95"
]
},
"execution_count": 57,
"execution_count": 59,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"\n",
"corr_scores"
"(corr_scores<pc_score).sum()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Corr grn from multiomics"
]
},
{
"cell_type": "code",
"execution_count": 58,
"execution_count": 69,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"0.6346976227062416"
]
},
"execution_count": 58,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"pc_score"
"multiomics_rna = ad.read_h5ad('resources/grn-benchmark/multiomics_rna.h5ad')"
]
},
{
"cell_type": "code",
"execution_count": 59,
"execution_count": 70,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"95"
]
},
"execution_count": 59,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"(corr_scores<pc_score).sum()"
"import scanpy as sc \n",
"sc.pp.normalize_total(multiomics_rna)\n",
"sc.pp.log1p(multiomics_rna)\n",
"sc.pp.scale(multiomics_rna)\n",
"X = multiomics_rna.X"
]
},
{
"cell_type": "code",
"execution_count": 60,
"execution_count": 75,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['AAGCCACGTCAGGCAT3', 'AAGGATGTCAGGCTAT3', 'AAGTTACGTTAGGCTA3',\n",
" 'ACCTGGATCCCTCAGT3', 'ACCTTGTGTACTTCAC3', 'AGACTATGTGCGCGTA3',\n",
" 'AGAGAGGAGCTTAACA3', 'AGCTATATCATGTCAA3', 'AGCTTGGTCCTAGTTT3',\n",
" 'AGTAGCTTCCTTAGGG3',\n",
" ...\n",
" 'CTTACTAGTGCCTCAC3', 'AACTAGCTCTGGCAAT3', 'TCCTCTAAGGCGCTAC3',\n",
" 'TCATGTTTCGTTACTT3', 'ACTTGAATCAGCACGC3', 'TAGTGTGGTTGGTTGA3',\n",
" 'GTCATGCCACCTCGCT3', 'AGTAAACAGTCACCAG3', 'ACCTCACTCAAGGACA3',\n",
" 'CTTAACAAGAGCCGGA3'],\n",
" dtype='object', length=2000)"
]
},
"execution_count": 60,
"metadata": {},
"output_type": "execute_result"
"name": "stderr",
"output_type": "stream",
"text": [
"Processing groups: 100%|██████████| 4/4 [00:43<00:00, 10.93s/it]\n"
]
}
],
"source": [
"import pandas as pd \n",
"pd.read_csv('resources_test/prior/cell_topic.csv', index_col=0).index"
"def corr_grn(X: np.ndarray, groups: np.ndarray):\n",
" grns = []\n",
" for group in tqdm(np.unique(groups), desc=\"Processing groups\"):\n",
" X_sub = X[groups == group, :]\n",
" X_sub = StandardScaler().fit_transform(X_sub)\n",
" grn = np.dot(X_sub.T, X_sub) / X_sub.shape[0]\n",
" grns.append(grn)\n",
" return np.mean(grns, axis=0)\n",
"groups = multiomics_rna.obs.cell_type\n",
"corr_net = corr_grn(X, groups)"
]
},
{
"cell_type": "code",
"execution_count": 61,
"execution_count": 79,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['GTCTTTAGTTAGTGAT3', 'TAATGCATCGAGGAGT3', 'TCTAGCACACTATGGC3',\n",
" 'GCTAATATCGTCAAGT3', 'GTTACGTAGGTGCGGA3', 'CATTCCTCATTAAGTC3',\n",
" 'GGGTGTTGTTACTTGC3', 'GCAAGCCTCTAAGTCA3', 'CAATCTAAGGCTACAT3',\n",
" 'AAGCGAGGTTGTTGCT3',\n",
" ...\n",
" 'GTGCTTACATGGCCCA3', 'ACGTTACAGCGAGCGA3', 'TGAGCACGTTGCCTCA3',\n",
" 'CTTAATGAGTTTCCTG3', 'AGGTGAATCGAGGAAC3', 'TCATGAGGTTTGTTGC3',\n",
" 'AAGTGAAGTCTTACTA3', 'ATACCTGCAGGAACTG3', 'CGCAATAGTGTAATAC3',\n",
" 'CTGCAATAGGTCCAAT3'],\n",
" dtype='object', name='obs_id', length=2000)"
]
},
"execution_count": 61,
"metadata": {},
"output_type": "execute_result"
}
],
"outputs": [],
"source": [
"ad.read_h5ad('resources_test/grn-benchmark/multiomics_rna.h5ad').obs.index.str.replace('-','')"
"corr_net = pd.DataFrame(corr_net, index=multiomics_rna.var_names, columns=multiomics_rna.var_names)"
]
},
{
"cell_type": "code",
"execution_count": 62,
"execution_count": 99,
"metadata": {},
"outputs": [],
"source": [
"tfs = corr_net.abs().sum(axis=0).argsort()[::-1][:1000].index.to_numpy()\n",
"corr_net_sub = corr_net[tfs]"
]
},
{
"cell_type": "code",
"execution_count": 100,
"metadata": {},
"outputs": [],
"source": [
"corr_net_sub = corr_net_sub.reset_index().melt(id_vars='location', var_name='source', value_name='weight')\n",
"corr_net_sub.rename(columns={'location': 'target'}, inplace=True)\n",
"corr_net_sub.to_csv('output/causal/grns/corr_net_sub.csv')"
]
},
{
"cell_type": "code",
"execution_count": 101,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Reading input files\n",
"(549, 15215)\n",
"Compute metrics for layer: pearson\n",
"ex(False)_tf(-1)\n",
"\n",
"Processing groups: 0%| | 0/5 [00:00<?, ?it/s]\n",
"Processing groups: 20%|██ | 1/5 [00:00<00:01, 2.06it/s]\n",
"Processing groups: 40%|████ | 2/5 [00:01<00:01, 1.99it/s]\n",
"Processing groups: 60%|██████ | 3/5 [00:01<00:00, 2.38it/s]\n",
"Processing groups: 80%|████████ | 4/5 [00:01<00:00, 2.28it/s]\n",
"Processing groups: 100%|██████████| 5/5 [00:02<00:00, 2.48it/s]\n",
"Processing groups: 100%|██████████| 5/5 [00:02<00:00, 2.35it/s]\n",
"/viash_automount/tmp/viash-run-regression_1-Itbpvt.py:48: FutureWarning: X.dtype being converted to np.float32 from float64. In the next version of anndata (0.9) conversion will not be automatic. Pass dtype explicitly to avoid this warning. Pass `AnnData(X, dtype=X.dtype, ...)` to get the future behavour.\n",
" output = ad.AnnData(\n",
" ex(False)_tf(-1) Mean\n",
"0 0.118948 0.118948\n",
"(2,) (2,)\n"
]
},
{
"data": {
"text/plain": [
"2000"
"CompletedProcess(args='viash run src/metrics/regression_1/config.vsh.yaml -- --prediction output/causal/grns/corr_net_sub.csv --score output/causal/scores/corr_net_sub.h5ad', returncode=0)"
]
},
"execution_count": 62,
"execution_count": 101,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import anndata as ad \n",
"ad.read_h5ad('resources_test/grn-benchmark/multiomics_rna.h5ad').obs.index.str.replace('-','').isin(pd.read_csv('resources_test/prior/cell_topic.csv', index_col=0).index).sum()"
]
},
{
"cell_type": "code",
"execution_count": 63,
"metadata": {},
"outputs": [],
"source": [
"cell_topic=pd.read_csv('output/scenicplus/cell_topic.csv', index_col=0)\n",
"cell_topic.index = cell_topic.index.str.split('-').str[0]\n",
"cell_topic.to_csv('output/scenicplus/cell_topic.csv')"
"cmd = f\"viash run src/metrics/regression_1/config.vsh.yaml -- --prediction output/causal/grns/corr_net_sub.csv --score output/causal/scores/corr_net_sub.h5ad\"\n",
"subprocess.run(cmd, shell=True)"
]
}
],
Expand Down
Loading

0 comments on commit 3282935

Please sign in to comment.