TCGA-BRCA Demo

Dataset Source

Omics Data: FireHose BRCA
Clinical and PAM50 Data: TCGAbiolinks

Dataset Overview

Original Data:

Methylation: 20,107 × 885
mRNA: 18,321 × 1,212
miRNA: 503 × 1,189
PAM50: 1,087 × 1
Clinical: 1,098 × 101
Note: Omics matrices are features × samples; clinical matrices are samples × fields.

PAM50 Subtype Counts (Original)

LumA: 419
LumB: 140
Basal: 130
Her2: 46
Normal: 34

Patients in Every Dataset

Total patients present in methylation, mRNA, miRNA, PAM50, and clinical: 769

Final Shapes (Per-Patient)

After aggregating multiple aliquots by mean, all modalities align on 769 patients:

Methylation: 769 × 20,107
mRNA: 769 × 20,531
miRNA: 769 × 503
PAM50: 769 × 1
Clinical: 769 × 119

Data Summary Table

Stage	Clinical	Methylation	miRNA	mRNA	PAM50 (Subtype Counts)	Notes
Original Raw Data	1,098 × 101	20,107 × 885	503 × 1,189	18,321 × 1,212	LumA: 509LumB: 209Basal: 192Her2: 82Normal: 40	Raw FireHose & TCGAbiolinks files
Patient-Level Intersection	769 × 101	769 × 20,107	769 × 1,046	769 × 20,531	LumA: 419LumB: 140Basal: 130Her2: 46Normal: 34	Patients with complete data in all sets

Reference Links

Lets take a look at the data from FireHose directly after download

[1]:

import pandas as pd
from pathlib import Path
root = Path("/home/vicente/Github/BioNeuralNet/TCGA_BRCA_DATA")

mirna_raw = pd.read_csv(root/"BRCA.miRseq_RPKM_log2.txt", sep="\t",index_col=0,low_memory=False)
rna_raw = pd.read_csv(root / "BRCA.uncv2.mRNAseq_RSEM_normalized_log2.txt", sep="\t",index_col=0,low_memory=False)
meth_raw = pd.read_csv(root/"BRCA.meth.by_mean.data.txt", sep='\t',index_col=0,low_memory=False)
clinical_raw = pd.read_csv(root / "BRCA.clin.merged.picked.txt",sep="\t", index_col=0, low_memory=False)

print(f"mirna shape: {mirna_raw.shape}, rna shape: {rna_raw.shape}, meth shape: {meth_raw.shape}, clinical shape: {clinical_raw.shape}")
display(mirna_raw.head())
display(rna_raw.head())
display(meth_raw.head())
display(clinical_raw.head())

mirna shape: (503, 1189), rna shape: (18321, 1212), meth shape: (20107, 885), clinical shape: (18, 1097)

	TCGA-3C-AAAU-01	TCGA-3C-AALI-01	TCGA-3C-AALJ-01	TCGA-3C-AALK-01	TCGA-4H-AAAK-01	TCGA-5L-AAT0-01	TCGA-5L-AAT1-01	TCGA-5T-A9QA-01	TCGA-A1-A0SB-01	TCGA-A1-A0SD-01	...	TCGA-BH-A0WA-01	TCGA-E2-A105-01	TCGA-E2-A106-01	TCGA-E2-A107-01	TCGA-E2-A108-01	TCGA-E2-A109-01	TCGA-E2-A10B-01	TCGA-E2-A10C-01	TCGA-E2-A10E-01	TCGA-E2-A10F-01
gene
hsa-let-7a-1	13.129765	12.918069	13.012033	13.144697	13.411684	13.316301	13.445230	13.727850	13.601504	13.598739	...	12.225132	13.938134	13.609853	13.508290	13.406359	13.730647	13.198426	12.793350	14.060268	12.990403
hsa-let-7a-2	14.117933	13.922300	14.010002	14.141721	14.413518	14.310917	14.448556	14.714551	14.608693	14.606942	...	13.235065	14.930021	14.603389	14.525026	14.402735	14.719166	14.200523	13.796623	15.047592	14.006035
hsa-let-7a-3	13.147714	12.913194	13.028483	13.151281	13.420481	13.327144	13.446806	13.736891	13.613105	13.606224	...	12.261971	13.972011	13.643274	13.549981	13.438737	13.732070	13.212367	12.793350	14.074978	13.018659
hsa-let-7b	14.595135	14.512657	13.419612	14.667196	14.438548	14.576493	14.611137	15.098805	16.505758	15.638855	...	14.684912	15.230457	15.357655	15.112011	15.040315	15.806771	15.645910	14.724106	16.370741	15.439239
hsa-let-7c	8.414890	9.646536	9.312455	11.511431	11.693927	11.138419	11.284446	9.197514	13.392164	11.419823	...	10.565698	10.483745	11.159056	12.473340	12.405828	10.613712	11.395452	9.087202	10.885520	11.385638

5 rows × 1189 columns

	TCGA-3C-AAAU-01	TCGA-3C-AALI-01	TCGA-3C-AALJ-01	TCGA-3C-AALK-01	TCGA-4H-AAAK-01	TCGA-5L-AAT0-01	TCGA-5L-AAT1-01	TCGA-5T-A9QA-01	TCGA-A1-A0SB-01	TCGA-A1-A0SD-01	...	TCGA-UL-AAZ6-01	TCGA-UU-A93S-01	TCGA-V7-A7HQ-01	TCGA-W8-A86G-01	TCGA-WT-AB41-01	TCGA-WT-AB44-01	TCGA-XX-A899-01	TCGA-XX-A89A-01	TCGA-Z7-A8R5-01	TCGA-Z7-A8R6-01
gene
?\|100133144	4.032489	3.211931	3.538886	3.595671	2.775430	1.995991	NaN	0.550310	3.939189	3.250628	...	-1.324816	2.108558	NaN	2.475707	NaN	NaN	3.846574	4.480524	1.178747	2.783771
?\|100134869	3.692829	4.119273	3.206237	3.469873	3.850979	3.766489	3.405298	3.169252	3.847346	3.501324	...	3.845189	3.443978	1.622556	3.845099	2.657434	1.703987	4.422294	4.769476	2.866572	4.631075
?\|10357	5.704604	6.124231	7.269570	7.168565	6.395968	6.836141	6.857961	6.749035	6.862786	5.913201	...	7.083470	7.088829	4.906766	7.003547	5.744909	5.401368	7.106177	6.003213	6.410173	7.388457
?\|10431	8.672694	9.139279	10.410275	9.757450	9.581922	9.657753	10.114256	10.472185	9.360367	9.933569	...	10.616682	11.495054	10.749770	9.446410	10.282241	10.874534	9.350400	9.497295	10.155173	9.970921
?\|155060	10.213110	9.011343	9.209506	9.110487	8.027083	8.110023	7.704865	6.254741	8.128052	6.387132	...	8.052478	7.516236	9.280761	9.631306	8.137225	9.460539	8.738651	8.556414	7.977670	7.894918

5 rows × 1212 columns

	TCGA-3C-AAAU-01	TCGA-3C-AALI-01	TCGA-3C-AALJ-01	TCGA-3C-AALK-01	TCGA-4H-AAAK-01	TCGA-5L-AAT0-01	TCGA-5L-AAT1-01	TCGA-5T-A9QA-01	TCGA-A1-A0SB-01	TCGA-A1-A0SE-01	...	TCGA-UL-AAZ6-01	TCGA-UU-A93S-01	TCGA-V7-A7HQ-01	TCGA-W8-A86G-01	TCGA-WT-AB41-01	TCGA-WT-AB44-01	TCGA-XX-A899-01	TCGA-XX-A89A-01	TCGA-Z7-A8R5-01	TCGA-Z7-A8R6-01
Hybridization REF
Composite Element REF	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	...	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value	Beta_Value
A1BG	0.483716119676	0.637191226131	0.656092398242	0.615194471357	0.612080370511	0.469600740678	0.582188239422	0.66617073097	0.659965611959	0.641701155202	...	0.631413241724	0.64952294395	0.596585169597	0.615558357651	0.580837880262	0.615814023324	0.589897794957	0.572606636128	0.617859586161	0.568150149265
A1CF	0.295827203492	0.458972998571	0.489725289638	0.625765223243	0.507736509665	0.514770866326	0.549850958729	0.381038654448	0.826312156393	0.606699429409	...	0.383469192855	0.183354853938	0.403909161312	0.716980255014	0.613131295074	0.665043713213	0.705153725375	0.494848686021	0.691835387189	0.224696596211
A2BP1	0.187699869591	0.240515847704	0.279087851226	0.488888510474	0.463845494635	0.504450855353	0.480885816745	0.622832399216	0.474678831563	0.339829506578	...	0.130529915536	0.319855310743	0.335517456053	0.512185396638	0.563519806811	0.507364324635	0.520542747167	0.412562068574	0.522169978143	0.33955834608
A2LD1	0.62958551322	0.666272288675	0.755630499986	0.74575121287	0.698515739124	0.706812706661	0.759017355996	0.694010939885	0.847837522256	0.786662091353	...	0.587475995313	0.667969642321	0.689140211036	0.791381283524	0.680499323148	0.660476360054	0.745725420412	0.74390049875	0.791229999577	0.637764188841

5 rows × 885 columns

	tcga-5l-aat0	tcga-5l-aat1	tcga-a1-a0sp	tcga-a2-a04v	tcga-a2-a04y	tcga-a2-a0cq	tcga-a2-a1g4	tcga-a2-a25a	tcga-a7-a0cd	tcga-a7-a13g	...	tcga-s3-aa11	tcga-s3-aa14	tcga-s3-aa15	tcga-ul-aaz6	tcga-uu-a93s	tcga-v7-a7hq	tcga-wt-ab44	tcga-xx-a899	tcga-xx-a89a	tcga-z7-a8r6
Hybridization REF
Composite Element REF	value	value	value	value	value	value	value	value	value	value	...	value	value	value	value	value	value	value	value	value	value
years_to_birth	42	63	40	39	53	62	71	44	66	79	...	67	47	51	73	63	75	NaN	46	68	46
vital_status	0	0	0	1	0	0	0	0	0	0	...	0	0	0	0	1	0	0	0	0	0
days_to_death	NaN	NaN	NaN	1920	NaN	NaN	NaN	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	116	NaN	NaN	NaN	NaN	NaN
days_to_last_followup	1477	1471	584	NaN	1099	2695	595	3276	1165	718	...	421	529	525	518	NaN	2033	883	467	488	3256

5 rows × 1097 columns

TCGAbiolinks

This section demonstrates how to use the TCGAbiolinks R package to access and download clinical and molecular subtype data. It begins by ensuring TCGAbiolinks is installed, then loads the package. It retrieves PAM50 molecular subtype labels using TCGAquery_subtype() and writes them to a CSV file. Additionally, it downloads clinical data using GDCquery_clinic() and formats it with GDCprepare_clinic(), saving the result as another CSV file.

# Install TCGAbiolinks
if (!requireNamespace("TCGAbiolinks", quietly = TRUE)) {
  if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")
  BiocManager::install("TCGAbiolinks")
}

# Load the library
library(TCGAbiolinks)

# Download PAM50 subtype labels
pam50_df <- TCGAquery_subtype(tumor = "BRCA")[ , c("patient", "BRCA_Subtype_PAM50")]
write.csv(pam50_df, file = "BRCA_PAM50_labels.csv", row.names = FALSE, quote = FALSE)

# Download clinical data
clin_raw <- GDCquery_clinic(project = "TCGA-BRCA", type = "clinical")
clin_df <- GDCprepare_clinic(clin_raw, clinical.info = "patient")
write.csv(clin_df, file = "BRCA_clinical_data.csv", row.names = FALSE, quote = FALSE)

[2]:

import pandas as pd

# from Firehose
mirna = pd.read_csv(root/"BRCA.miRseq_RPKM_log2.txt", sep="\t",index_col=0,low_memory=False)
meth = pd.read_csv(root/"BRCA.meth.by_mean.data.txt", sep='\t',index_col=0,low_memory=False)
rna = pd.read_csv(root / "BRCA.uncv2.mRNAseq_RSEM_normalized_log2.txt", sep="\t",index_col=0,low_memory=False)
clinical_firehose = pd.read_csv(root / "BRCA.clin.merged.picked.txt",sep="\t", index_col=0, low_memory=False).T

# from TCGABiolinks
pam50 = pd.read_csv(root /"BRCA_PAM50_labels.csv",index_col=0)
clinical_biolinks = pd.read_csv(root /"BRCA_clinical_data.csv",index_col=1)

print("Initial shapes")
print(f"meth: {meth.shape}")
print(f"rna: {rna.shape}")
print(f"mirna: {mirna.shape}")
print(f"pam50: {pam50.shape}")
print(f"clinical TCGABioLinks: {clinical_biolinks.shape}")
print(f"clinical FireHose: {clinical_firehose.shape}")

meth = meth.T
rna = rna.T
mirna = mirna.T

print("\nAfter tranpose")
print(f"meth: {meth.shape}")
print(f"rna: {rna.shape}")
print(f"mirna: {mirna.shape}")

def trim(idx):
    return idx.to_series().str.extract(r'(^TCGA-\w\w-\w\w\w\w)')[0]

meth.index = trim(meth.index)
rna.index = trim(rna.index)
mirna.index = trim(mirna.index)
pam50.index = pam50.index.str.upper()
clinical_biolinks.index = clinical_biolinks.index.str.upper()
clinical_firehose.index = clinical_firehose.index.str.upper()

idx1 = clinical_biolinks.index
idx2 = clinical_firehose.index

# intersection and unique counts
common = idx1.intersection(idx2)
only_in_1 = idx1.difference(idx2)
only_in_2 = idx2.difference(idx1)

print(f"Patients in both clinical datasets: {len(common)}")
common = clinical_biolinks.index.intersection(clinical_firehose.index)
clinical_biolinks = clinical_biolinks.loc[common]
clinical_firehose = clinical_firehose.loc[common]

clinical = pd.concat([clinical_biolinks, clinical_firehose], axis=1)

print(f"Combined Clinical shape {clinical.shape}")

common = sorted(set(meth.index) & set(rna.index) & set(mirna.index) & set(pam50.index) & set(clinical.index))
print(f"Patients in every dataset: {len(common)}")

meth = meth.loc[common]
rna = rna.loc[common]
mirna = mirna.loc[common]
pam50 = pam50.loc[common]
clinical = clinical.loc[common]

print("\nFinal shapes:")
print(f"meth: {meth.shape}")
print(f"rna: {rna.shape}")
print(f"mirna: {mirna.shape}")
print(f"pam50: {pam50.shape}")
print(f"clinical: {clinical.shape}\n")

Initial shapes
meth: (20107, 885)
rna: (18321, 1212)
mirna: (503, 1189)
pam50: (1087, 1)
clinical TCGABioLinks: (1098, 101)
clinical FireHose: (1097, 18)

After tranpose
meth: (885, 20107)
rna: (1212, 18321)
mirna: (1189, 503)
Patients in both clinical datasets: 1097
Combined Clinical shape (1097, 119)
Patients in every dataset: 769

Final shapes:
meth: (863, 20107)
rna: (865, 18321)
mirna: (855, 503)
pam50: (769, 1)
clinical: (769, 119)

Handling Multiple Aliquots per Sample

This section addresses cases where some patients have multiple aliquots per sample in the meth, rna, and mirna datasets. It first identifies and counts patients with duplicate entries. Then, it coerces all data to numeric types and aggregates the duplicates by computing the mean across aliquots for each patient, ensuring only one row per patient. After aggregation, the datasets are aligned by keeping only the patients that are common across all five datasets (meth, rna, mirna, pam50, and clinical). The result is s set of matched samples ready for integrated analysis.

[3]:

for name, df in [("meth", meth), ("rna", rna), ("mirna", mirna)]:
    counts = df.index.value_counts()
    n_multiple = (counts > 1).sum()
    total_duplicates = counts[counts > 1].sum() - n_multiple

    print(f"{name}:")
    print(f"patients with >1 aliquot: {n_multiple}")
    print(f"total duplicate rows: {total_duplicates}\n")

meth = meth.apply(pd.to_numeric, errors="coerce")
rna = rna .apply(pd.to_numeric, errors="coerce")
mirna = mirna.apply(pd.to_numeric, errors="coerce")

meth = meth.groupby(level=0).mean()
rna = rna.groupby(level=0).mean()
mirna = mirna.groupby(level=0).mean()

# Now each has one row per patient
print("Post-aggregation shapes:")
print(f"meth: {meth.shape}")
print(f"rna: {rna.shape}")
print(f"mirna: {mirna.shape}")

common = sorted( set(meth.index) & set(rna.index) & set(mirna.index)& set(pam50.index) & set(clinical.index) )
print(f"Patients in every dataset: {len(common)}")

meth = meth.loc[common]
rna = rna.loc[common]
mirna = mirna.loc[common]
pam50 = pam50.loc[common]
clinical = clinical.loc[common]

print("\nFinal shapes")
print(f"meth: {meth.shape}")
print(f"rna: {rna.shape}")
print(f"mirna: {mirna.shape}")
print(f"pam50: {pam50.shape}")
print(f"clinical:{clinical.shape}")

meth:
patients with >1 aliquot: 91
total duplicate rows: 94

rna:
patients with >1 aliquot: 93
total duplicate rows: 96

mirna:
patients with >1 aliquot: 84
total duplicate rows: 86

Post-aggregation shapes:
meth: (769, 20107)
rna: (769, 18321)
mirna: (769, 503)
Patients in every dataset: 769

Final shapes
meth: (769, 20107)
rna: (769, 18321)
mirna: (769, 503)
pam50: (769, 1)
clinical:(769, 119)

Review the first few rows of each file

[4]:

display(meth.head())
display(rna.head())
display(mirna.head())
display(clinical.head())
display(pam50.value_counts())

Hybridization REF	Composite Element REF	A1BG	A1CF	A2BP1	A2LD1	A2M	A2ML1	A4GALT	A4GNT	AAA1	...	ZWILCH	ZWINT	ZXDC	ZYG11A	ZYG11B	ZYX	ZZEF1	ZZZ3	psiTPTE22	tAKR
0
TCGA-3C-AAAU	NaN	0.483716	0.295827	0.187700	0.629586	0.559654	0.835412	0.484800	0.690217	0.807805	...	0.112978	0.053939	0.287665	0.328087	0.502935	0.220683	0.482044	0.107396	0.247304	0.506404
TCGA-3C-AALI	NaN	0.637191	0.458973	0.240516	0.666272	0.607505	0.842391	0.550047	0.749890	0.395290	...	0.111834	0.046160	0.265322	0.405851	0.434024	0.236362	0.458847	0.119652	0.163022	0.623865
TCGA-3C-AALJ	NaN	0.656092	0.489725	0.279088	0.755630	0.662360	0.829020	0.476107	0.653756	0.795102	...	0.113218	0.042657	0.272103	0.391326	0.449525	0.210976	0.482641	0.102385	0.252328	0.504451
TCGA-3C-AALK	NaN	0.615194	0.625765	0.488889	0.745751	0.727982	0.835365	0.556016	0.652005	0.816423	...	0.145133	0.047022	0.301284	0.410348	0.446571	0.220185	0.485944	0.112941	0.471956	0.682468
TCGA-4H-AAAK	NaN	0.612080	0.507737	0.463845	0.698516	0.692364	0.802388	0.504870	0.531183	0.851114	...	0.118928	0.045057	0.300647	0.379998	0.487929	0.233324	0.490736	0.115646	0.314877	0.744877

5 rows × 20107 columns

gene	?\|100133144	?\|100134869	?\|10357	?\|10431	?\|155060	?\|26823	?\|340602	?\|388795	?\|390284	?\|391343	...	ZWINT\|11130	ZXDA\|7789	ZXDB\|158586	ZXDC\|79364	ZYG11A\|440590	ZYG11B\|79699	ZYX\|7791	ZZEF1\|23140	ZZZ3\|26009	psiTPTE22\|387590
0
TCGA-3C-AAAU	4.032489	3.692829	5.704604	8.672694	10.213110	NaN	0.785174	-1.536587	2.048201	NaN	...	9.864120	7.017830	9.976968	10.695662	8.013988	10.238851	11.776124	10.887932	10.205129	0.785174
TCGA-3C-AALI	3.211931	4.119273	6.124231	9.139279	9.011343	0.121015	7.170928	2.291014	0.706022	3.027968	...	9.914682	5.902438	8.809329	10.391374	7.632831	9.237422	12.426428	10.364848	8.667973	9.855788
TCGA-3C-AALJ	3.538886	3.206237	7.269570	10.410275	9.209506	NaN	NaN	1.443554	1.443554	NaN	...	11.305650	5.143969	9.060691	9.586488	8.374267	9.055784	12.414355	9.880935	8.992994	5.143969
TCGA-3C-AALK	3.595671	3.469873	7.168565	9.757450	9.110487	-1.273343	NaN	1.048724	2.186215	NaN	...	9.384994	5.782065	8.773906	9.754688	7.454703	9.246419	12.474556	9.609426	9.453001	6.057699
TCGA-4H-AAAK	2.775430	3.850979	6.395968	9.581922	8.027083	-1.232769	-1.232769	1.574683	1.574683	NaN	...	9.397606	5.612830	8.728789	10.035881	3.811738	9.599438	11.980747	9.700292	9.784147	7.548699

5 rows × 18321 columns

gene	hsa-let-7a-1	hsa-let-7a-2	hsa-let-7a-3	hsa-let-7b	hsa-let-7c	hsa-let-7d	hsa-let-7e	hsa-let-7f-1	hsa-let-7f-2	hsa-let-7g	...	hsa-mir-937	hsa-mir-939	hsa-mir-940	hsa-mir-942	hsa-mir-944	hsa-mir-95	hsa-mir-96	hsa-mir-98	hsa-mir-99a	hsa-mir-99b
0
TCGA-3C-AAAU	13.129765	14.117933	13.147714	14.595135	8.414890	8.665921	10.521777	3.879392	11.824817	8.597744	...	0.906699	-0.093302	2.672234	2.467414	1.044202	2.044202	6.906699	5.754696	7.024602	15.506461
TCGA-3C-AALI	12.918069	13.922300	12.913194	14.512657	9.646536	9.003653	9.131760	4.386952	12.678841	8.455144	...	1.579597	-0.083367	0.139024	3.032109	-0.668331	0.331670	5.912870	6.427066	7.885299	13.626182
TCGA-3C-AALJ	13.012033	14.010002	13.028483	13.419612	9.312455	9.276943	11.395711	5.314692	13.530255	9.230563	...	3.270298	-2.189134	0.395828	1.855261	-0.381778	0.717757	6.603657	6.878301	7.580704	15.013822
TCGA-3C-AALK	13.144697	14.141721	13.151281	14.667196	11.511431	8.384763	10.368981	4.159182	12.652559	8.471503	...	0.923965	-0.660997	-0.076034	1.798435	1.798435	0.798435	6.181354	5.377922	10.031619	14.554783
TCGA-4H-AAAK	13.411684	14.413518	13.420481	14.438548	11.693927	8.453747	10.741371	4.494537	13.009499	8.381220	...	0.182950	-0.624403	-1.624403	1.076036	0.182950	-0.302475	4.318110	5.103516	10.078201	14.650338

5 rows × 503 columns

	project	synchronous_malignancy	ajcc_pathologic_stage	laterality	created_datetime	last_known_disease_status	tissue_or_organ_of_origin	days_to_last_follow_up	age_at_diagnosis	...	pathology_N_stage	pathology_M_stage	gender	date_of_initial_pathologic_diagnosis	days_to_last_known_alive	radiation_therapy	histological_type	number_of_lymph_nodes	race	ethnicity
TCGA-3C-AAAU	TCGA-BRCA	No	Stage X	Left	NaN	NaN	Breast, NOS	NaN	20211.0	...	nx	mx	female	2004	NaN	no	infiltrating lobular carcinoma	4	white	not hispanic or latino
TCGA-3C-AALI	TCGA-BRCA	No	Stage IIB	Right	NaN	NaN	Breast, NOS	NaN	18538.0	...	n1a	m0	female	2003	NaN	yes	infiltrating ductal carcinoma	1	black or african american	not hispanic or latino
TCGA-3C-AALJ	TCGA-BRCA	No	Stage IIB	Right	NaN	NaN	Breast, NOS	NaN	22848.0	...	n1a	m0	female	2011	NaN	no	infiltrating ductal carcinoma	1	black or african american	not hispanic or latino
TCGA-3C-AALK	TCGA-BRCA	No	Stage IA	Right	NaN	NaN	Breast, NOS	NaN	19074.0	...	n0 (i+)	m0	female	2011	NaN	no	infiltrating ductal carcinoma	0	black or african american	not hispanic or latino
TCGA-4H-AAAK	TCGA-BRCA	No	Stage IIIA	Left	NaN	NaN	Breast, NOS	NaN	18371.0	...	n2a	m0	female	2013	NaN	no	infiltrating lobular carcinoma	4	white	not hispanic or latino

5 rows × 119 columns

BRCA_Subtype_PAM50
LumA                  419
LumB                  140
Basal                 130
Her2                   46
Normal                 34
Name: count, dtype: int64

Preprocessing

After reviewing the data above, we applied the following steps to the data before further analysis.

Methylation (B -> M-value)
- Clip B-values to [E, 1-E] and apply logit transform: M = log_2(B / (1-B)).
- Drop the original Composite Element REF column.
mRNA & miRNA:
- Already in log_2 scale (RSEM normalized and RPKM).
Quality Control:
- Count samples with all-zero rows in each modality.
- Compute NaN counts post-transformation, then replace all NaNs with 0.
Column Name Cleaning:
- Replace all - and | characters with _.
- Replace ? with unknown.
Label Encoding:
- Map PAM50 subtypes to integers: Normal=0, Basal=1, Her2=2, LumA=3, LumB=4
Alignment & Aggregation:
- Trim barcodes to patient level.
- Aggregate duplicate aliquots by mean per patient.
- Drop the project column from clinical.
- Subset all tables to the common patient set (no missing or all-zero samples).
Final Output Shapes:
- Methylation M-value: 769 × 20,107
- mRNA (log_2): 769 × 20,531
- miRNA (log_2): 769 × 503
- PAM50 labels: 769 × 1
- Clinical covariates: 769 × 101

[ ]:

import numpy as np
import pandas as pd

def beta_to_m(df, eps=1e-6):
    B = np.clip(df.values, eps, 1.0 - eps)
    M = np.log2(B / (1 - B))
    return pd.DataFrame(M, index=df.index, columns=df.columns)

# find rows that are all 0s
zeros_meth = (meth  == 0).all(axis=1).sum()
zeros_rna = (rna   == 0).all(axis=1).sum()
zeros_mirna = (mirna == 0).all(axis=1).sum()
print(f"All zeros: meth: {zeros_meth}, rna: {zeros_rna}, mirna: {zeros_mirna}")

# find rows with all nans
nan_meth = meth.isna().all(axis=1).sum()
nan_rna = rna.isna().all(axis=1).sum()
nan_mirna = mirna.isna().all(axis=1).sum()
nan_clinical = clinical.isna().all(axis=1).sum()
nan_pam50 = pam50.isna().all(axis=1).sum()
print(f"nan_meth: {nan_meth}, nan_rna: {nan_rna}, nan_mirna: {nan_mirna}, nan_clinical: {nan_clinical}, nan_pam50: {nan_pam50}")

# map PAM50 subtypes to integers
mapping = {"Normal":0, "Basal":1, "Her2":2, "LumA":3, "LumB":4}
pam50 = pam50["BRCA_Subtype_PAM50"].map(mapping).to_frame(name="pam50")

# drop and transform methylation
meth_clean = meth.drop(columns=["Composite Element REF"], errors="ignore")
meth_m = beta_to_m(meth_clean)
clinical = clinical.drop(columns=["project"], errors="ignore")

# clean column names and fill nans
for df in [meth_m, rna, mirna]:
    df.columns = df.columns.str.replace(r"\?\|", "unknown_", regex=True)
    df.columns = df.columns.str.replace(r"[?|]", "unknown_", regex=True)
    df.columns = df.columns.str.replace("-", "_", regex=False)
    df.columns = df.columns.str.replace(r"_+", "_", regex=True)
    df.fillna(0, inplace=True)

# check for nans after filling
print("NaN counts after filling:")
print(meth_m.isna().sum().sum(),rna.isna().sum().sum(),mirna.isna().sum().sum(),clinical.isna().sum().sum(),pam50.isna().sum().sum())

# align index to PAM50
X_meth = meth_m.loc[pam50.index]
X_rna = rna.loc[pam50.index]
X_mirna = mirna.loc[pam50.index]
clinical= clinical.loc[pam50.index]

print(f"new shapes: meth: {X_meth.shape}, rna: {X_rna.shape}, mirna: {X_mirna.shape}, pam50: {pam50.shape}, clinical: {clinical.shape}")
display(X_meth.head())
display(X_rna.head())
display(X_mirna.head())
display(clinical.head())
display(pam50.value_counts())

All zeros: meth: 0, rna: 0, mirna: 0
nan_meth: 0, nan_rna: 0, nan_mirna: 0, nan_clinical: 0, nan_pam50: 0
NaN counts after filling:
0 0 0 46476 0
new shapes: meth: (769, 20106), rna: (769, 18321), mirna: (769, 503), pam50: (769, 1), clinical: (769, 118)

Hybridization REF	A1BG	A1CF	A2BP1	A2LD1	A2M	A2ML1	A4GALT	A4GNT	AAA1	AAAS	...	ZWILCH	ZWINT	ZXDC	ZYG11A	ZYG11B	ZYX	ZZEF1	ZZZ3	psiTPTE22	tAKR
patient
TCGA-3C-AAAU	-0.094004	-1.251175	-2.113585	0.765262	0.345896	2.343631	-0.087741	1.155791	2.071436	-2.650851	...	-2.972923	-4.132523	-1.308165	-1.034199	0.016935	-1.820233	-0.103662	-3.055084	-1.605783	0.036955
TCGA-3C-AALI	0.812517	-0.237291	-1.658888	0.997440	0.630221	2.418135	0.289780	1.584114	-0.613329	-4.072465	...	-2.989465	-4.369032	-1.469365	-0.549876	-0.382967	-1.691887	-0.238022	-2.879231	-2.360128	0.729981
TCGA-3C-AALJ	0.931878	-0.059301	-1.369104	1.628617	0.972130	2.277584	-0.137988	0.916964	1.956230	-3.781647	...	-2.969472	-4.488190	-1.419578	-0.637297	-0.292273	-1.902991	-0.100215	-3.132087	-1.567104	0.025686
TCGA-3C-AALK	0.676913	0.741678	-0.064133	1.552454	1.420200	2.343133	0.324621	0.905816	2.152928	-3.894574	...	-2.558319	-4.341028	-1.213585	-0.523013	-0.309506	-1.824419	-0.081137	-2.973455	-0.162004	1.103860
TCGA-4H-AAAK	0.657963	0.044649	-0.209004	1.212210	1.170304	2.021628	0.028103	0.180184	2.515149	-3.885526	...	-2.889175	-4.405580	-1.217950	-0.706284	-0.069670	-1.716283	-0.053464	-2.934908	-1.121575	1.545812

5 rows × 20106 columns

gene	unknown_100133144	unknown_100134869	unknown_10357	unknown_10431	unknown_155060	unknown_26823	unknown_340602	unknown_388795	unknown_390284	unknown_391343	...	ZWINTunknown_11130	ZXDAunknown_7789	ZXDBunknown_158586	ZXDCunknown_79364	ZYG11Aunknown_440590	ZYG11Bunknown_79699	ZYXunknown_7791	ZZEF1unknown_23140	ZZZ3unknown_26009	psiTPTE22unknown_387590
patient
TCGA-3C-AAAU	4.032489	3.692829	5.704604	8.672694	10.213110	0.000000	0.785174	-1.536587	2.048201	0.000000	...	9.864120	7.017830	9.976968	10.695662	8.013988	10.238851	11.776124	10.887932	10.205129	0.785174
TCGA-3C-AALI	3.211931	4.119273	6.124231	9.139279	9.011343	0.121015	7.170928	2.291014	0.706022	3.027968	...	9.914682	5.902438	8.809329	10.391374	7.632831	9.237422	12.426428	10.364848	8.667973	9.855788
TCGA-3C-AALJ	3.538886	3.206237	7.269570	10.410275	9.209506	0.000000	0.000000	1.443554	1.443554	0.000000	...	11.305650	5.143969	9.060691	9.586488	8.374267	9.055784	12.414355	9.880935	8.992994	5.143969
TCGA-3C-AALK	3.595671	3.469873	7.168565	9.757450	9.110487	-1.273343	0.000000	1.048724	2.186215	0.000000	...	9.384994	5.782065	8.773906	9.754688	7.454703	9.246419	12.474556	9.609426	9.453001	6.057699
TCGA-4H-AAAK	2.775430	3.850979	6.395968	9.581922	8.027083	-1.232769	-1.232769	1.574683	1.574683	0.000000	...	9.397606	5.612830	8.728789	10.035881	3.811738	9.599438	11.980747	9.700292	9.784147	7.548699

5 rows × 18321 columns

gene	hsa_let_7a_1	hsa_let_7a_2	hsa_let_7a_3	hsa_let_7b	hsa_let_7c	hsa_let_7d	hsa_let_7e	hsa_let_7f_1	hsa_let_7f_2	hsa_let_7g	...	hsa_mir_937	hsa_mir_939	hsa_mir_940	hsa_mir_942	hsa_mir_944	hsa_mir_95	hsa_mir_96	hsa_mir_98	hsa_mir_99a	hsa_mir_99b
patient
TCGA-3C-AAAU	13.129765	14.117933	13.147714	14.595135	8.414890	8.665921	10.521777	3.879392	11.824817	8.597744	...	0.906699	-0.093302	2.672234	2.467414	1.044202	2.044202	6.906699	5.754696	7.024602	15.506461
TCGA-3C-AALI	12.918069	13.922300	12.913194	14.512657	9.646536	9.003653	9.131760	4.386952	12.678841	8.455144	...	1.579597	-0.083367	0.139024	3.032109	-0.668331	0.331670	5.912870	6.427066	7.885299	13.626182
TCGA-3C-AALJ	13.012033	14.010002	13.028483	13.419612	9.312455	9.276943	11.395711	5.314692	13.530255	9.230563	...	3.270298	-2.189134	0.395828	1.855261	-0.381778	0.717757	6.603657	6.878301	7.580704	15.013822
TCGA-3C-AALK	13.144697	14.141721	13.151281	14.667196	11.511431	8.384763	10.368981	4.159182	12.652559	8.471503	...	0.923965	-0.660997	-0.076034	1.798435	1.798435	0.798435	6.181354	5.377922	10.031619	14.554783
TCGA-4H-AAAK	13.411684	14.413518	13.420481	14.438548	11.693927	8.453747	10.741371	4.494537	13.009499	8.381220	...	0.182950	-0.624403	-1.624403	1.076036	0.182950	-0.302475	4.318110	5.103516	10.078201	14.650338

5 rows × 503 columns

	synchronous_malignancy	ajcc_pathologic_stage	days_to_diagnosis	laterality	created_datetime	last_known_disease_status	tissue_or_organ_of_origin	days_to_last_follow_up	age_at_diagnosis	primary_diagnosis	...	pathology_N_stage	pathology_M_stage	gender	date_of_initial_pathologic_diagnosis	days_to_last_known_alive	radiation_therapy	histological_type	number_of_lymph_nodes	race	ethnicity
patient
TCGA-3C-AAAU	No	Stage X	0.0	Left	NaN	NaN	Breast, NOS	NaN	20211.0	Lobular carcinoma, NOS	...	nx	mx	female	2004	NaN	no	infiltrating lobular carcinoma	4	white	not hispanic or latino
TCGA-3C-AALI	No	Stage IIB	0.0	Right	NaN	NaN	Breast, NOS	NaN	18538.0	Infiltrating duct carcinoma, NOS	...	n1a	m0	female	2003	NaN	yes	infiltrating ductal carcinoma	1	black or african american	not hispanic or latino
TCGA-3C-AALJ	No	Stage IIB	0.0	Right	NaN	NaN	Breast, NOS	NaN	22848.0	Infiltrating duct carcinoma, NOS	...	n1a	m0	female	2011	NaN	no	infiltrating ductal carcinoma	1	black or african american	not hispanic or latino
TCGA-3C-AALK	No	Stage IA	0.0	Right	NaN	NaN	Breast, NOS	NaN	19074.0	Infiltrating duct carcinoma, NOS	...	n0 (i+)	m0	female	2011	NaN	no	infiltrating ductal carcinoma	0	black or african american	not hispanic or latino
TCGA-4H-AAAK	No	Stage IIIA	0.0	Left	NaN	NaN	Breast, NOS	NaN	18371.0	Lobular carcinoma, NOS	...	n2a	m0	female	2013	NaN	no	infiltrating lobular carcinoma	4	white	not hispanic or latino

5 rows × 118 columns

pam50
3        419
4        140
1        130
2         46
0         34
Name: count, dtype: int64

[6]:

# lets set up a commong index for all the files and then save them to csv
X_meth.index.name = "patient"
X_rna.index.name = "patient"
X_mirna.index.name = "patient"
pam50.index.name = "patient"
clinical.index.name = "patient"

X_meth.to_csv(root / "meth.csv", index=True)
X_rna.to_csv(root / "rna.csv", index=True)
X_mirna.to_csv(root / "mirna.csv", index=True)
pam50.to_csv(root / "pam50.csv", index=True)
clinical.to_csv(root / "clinical.csv", index=True)

Optional: Load the data we just saved to make sure it looks okay.

[ ]:

meth = pd.read_csv(root / "meth.csv", index_col=0)
rna = pd.read_csv(root / "rna.csv", index_col=0)
mirna = pd.read_csv(root / "mirna.csv", index_col=0)
pam50 = pd.read_csv(root / "pam50.csv", index_col=0)
clinical = pd.read_csv(root / "clinical.csv", index_col=0)

display(meth.head())
display(rna.head())
display(mirna.head())
display(clinical.head())
display(pam50.head())

Easy Access via DatasetLoader

To facilitate working with this data, we have made it available through our DatasetLoader component. If you have additional pre-processed or raw datasets you would like to include, feel free to reach out and we are happy to support adding new datasets to the platform.

[8]:

from bioneuralnet.datasets import DatasetLoader

tgca_brca = DatasetLoader("brca")

print(f"TGCA BRCA dataset shape: {tgca_brca.shape}")
brca_meth = tgca_brca.data["meth"]
brca_rna = tgca_brca.data["rna"]
brca_mirna = tgca_brca.data["mirna"]
brca_clinical = tgca_brca.data["clinical"]
brca_pam50 = tgca_brca.data["pam50"]

TGCA BRCA dataset shape: {'mirna': (769, 503), 'pam50': (769, 1), 'clinical': (769, 118), 'meth': (769, 20106), 'rna': (769, 18321)}

[ ]:

from bioneuralnet.utils.preprocess import preprocess_clinical

#shapes
print(f"RNA shape: {brca_rna.shape}")
print(f"METH shape: {brca_meth.shape}")
print(f"miRNA shape: {brca_mirna.shape}")
print(f"Clinical shape: {brca_clinical.shape}")
print(f"Phenotype shape: {brca_pam50.shape}")
print(f"Phenotype counts:\n{brca_pam50.value_counts()}")

# review min and max values from the datasets
for name, df in {"rna": brca_rna, "meth": brca_meth, "mirna": brca_mirna}.items():
    min_val = df.min().min()
    max_val = df.max().max()
    print(f"\n{name.upper()}:")
    print(f"Min: {min_val:.4f}")
    print(f"Max: {max_val:.4f}")

#check nans in pam50
print(f"Nan values in pam50 {brca_pam50.isna().sum().sum()}")

brca_pam50 = brca_pam50.dropna()
X_rna = brca_rna.loc[brca_pam50.index]
X_meth = brca_meth.loc[brca_pam50.index]
X_mirna = brca_mirna.loc[brca_pam50.index]
clinical = brca_clinical.loc[brca_pam50.index]

# for more details on the preprocessing function, see bioneuralnet.utils.preprocess
clinical = preprocess_clinical(clinical, brca_pam50, top_k=15, scale=True, ignore_columns=["days_to_birth", "age_at_diagnosis", "days_to_last_followup", "age_at_index", "years_to_birth"])
display(clinical.head())

2025-05-16 10:31:09,364 - bioneuralnet.utils.preprocess - INFO - [Inf]: Replaced 0 infinite values
2025-05-16 10:31:09,365 - bioneuralnet.utils.preprocess - INFO - [NaN]: Replaced 31384 NaNs after median imputation
2025-05-16 10:31:09,365 - bioneuralnet.utils.preprocess - INFO - [Zero-Var]: 39 columns dropped due to zero variance

RNA shape: (769, 18321)
METH shape: (769, 20106)
miRNA shape: (769, 503)
Clinical shape: (769, 118)
Phenotype shape: (769, 1)
Phenotype counts:
pam50
3        419
4        140
1        130
2         46
0         34
Name: count, dtype: int64

RNA:
Min: -8.5873
Max: 20.9784

METH:
Min: -7.1642
Max: 6.9710

MIRNA:
Min: -4.4631
Max: 19.3838
Nan values in pam50 0

2025-05-16 10:31:09,752 - bioneuralnet.utils.preprocess - INFO - Selected top 15 features by RandomForest importance

	days_to_birth	age_at_diagnosis	days_to_last_followup	age_at_index	years_to_birth	year_of_diagnosis	number_of_lymph_nodes	date_of_initial_pathologic_diagnosis	histological_type_infiltrating lobular carcinoma	primary_diagnosis_Lobular carcinoma, NOS	morphology_8520/3	race.1_white	days_to_death.1	laterality_Right	primary_diagnosis_Infiltrating duct carcinoma, NOS	country_of_residence_at_enrollment_United States	sites_of_involvement_Breast, NOS	days_to_death	race_white	ajcc_staging_system_edition_6th
patient
TCGA-3C-AAAU	-20211.0	20211.0	4047.0	55.0	55.0	-1.50	1.5	-1.50	True	True	True	True	0.0	False	False	True	False	0.0	True	True
TCGA-3C-AALI	-18538.0	18538.0	4005.0	50.0	50.0	-1.75	0.0	-1.75	False	False	False	False	0.0	True	True	True	False	0.0	False	True
TCGA-3C-AALJ	-22848.0	22848.0	1474.0	62.0	62.0	0.25	0.0	0.25	False	False	False	False	0.0	True	True	True	True	0.0	False	False
TCGA-3C-AALK	-19074.0	19074.0	1448.0	52.0	52.0	0.25	-0.5	0.25	False	False	False	False	0.0	True	True	True	True	0.0	False	False
TCGA-4H-AAAK	-18371.0	18371.0	348.0	50.0	50.0	0.75	1.5	0.75	True	True	True	True	0.0	False	False	False	False	0.0	True	False

Preparing Multi-Omics Data for downstream tasks

Check sample overlap.
Select top features.
- Uses ANOVA F-test to select the most relevant features for classification from each omics dataset.
Combine datasets.
- Selected features from RNA, methylation, and miRNA are combined into a single dataset.
Clean missing values.
- Counts and removes any missing (nan) values from the combined dataset.
Build similarity graph.
- Creates a k-nearest neighbors graph from the transposed feature matrix.

Note: For more details on preprocessing functions and graph generation algorithms, see the Utils documentation

[ ]:

from sklearn.metrics import accuracy_score, f1_score
from bioneuralnet.utils.preprocess import top_anova_f_features
from bioneuralnet.utils.graph import gen_similarity_graph

#count intersection of samples
print("Intersection of samples:")
print(f"RNA: {len(set(X_rna.index) & set(pam50.index))}")
print(f"METH: {len(set(X_meth.index) & set(pam50.index))}")
print(f"miRNA: {len(set(X_mirna.index) & set(pam50.index))}")
print(f"Clinical: {len(set(clinical.index) & set(pam50.index))}")

meth_sel = top_anova_f_features(X_meth, brca_pam50, max_features=1000, task="classification")
rna_sel = top_anova_f_features(X_rna, brca_pam50 ,max_features=1000, task="classification")
mirna_sel = top_anova_f_features(X_mirna, brca_pam50,max_features=503, task="classification")
X_train_full = pd.concat([meth_sel, rna_sel, mirna_sel], axis=1)

#count nan values
print(f"Nan values in X_train_full: {X_train_full.isna().sum().sum()}")

#drop nan values
X_train_full = X_train_full.dropna()

#check if there are any nan values
print(f"Nan value in X_train_full after dropping: {X_train_full.isna().sum().sum()}")

print(f"X_train_full shape: {X_train_full.shape}")
A_train = gen_similarity_graph(X_train_full.T, k=15)

print(f"Network shape: {A_train.shape}")

Intersection of samples:
RNA: 769
METH: 769
miRNA: 769
Clinical: 769

2025-05-16 10:31:12,677 - bioneuralnet.utils.preprocess - INFO - [Inf]: Replaced 0 infinite values
2025-05-16 10:31:12,678 - bioneuralnet.utils.preprocess - INFO - [NaN]: Replaced 0 NaNs after median imputation
2025-05-16 10:31:12,678 - bioneuralnet.utils.preprocess - INFO - [Zero-Var]: 0 columns dropped due to zero variance
2025-05-16 10:31:12,835 - bioneuralnet.utils.preprocess - INFO - Selected 1000 features by ANOVA (task=classification), 17514 significant, 0 padded
2025-05-16 10:31:15,470 - bioneuralnet.utils.preprocess - INFO - [Inf]: Replaced 0 infinite values
2025-05-16 10:31:15,471 - bioneuralnet.utils.preprocess - INFO - [NaN]: Replaced 0 NaNs after median imputation
2025-05-16 10:31:15,471 - bioneuralnet.utils.preprocess - INFO - [Zero-Var]: 0 columns dropped due to zero variance
2025-05-16 10:31:15,635 - bioneuralnet.utils.preprocess - INFO - Selected 1000 features by ANOVA (task=classification), 16864 significant, 0 padded
2025-05-16 10:31:15,714 - bioneuralnet.utils.preprocess - INFO - [Inf]: Replaced 0 infinite values
2025-05-16 10:31:15,715 - bioneuralnet.utils.preprocess - INFO - [NaN]: Replaced 0 NaNs after median imputation
2025-05-16 10:31:15,715 - bioneuralnet.utils.preprocess - INFO - [Zero-Var]: 0 columns dropped due to zero variance
2025-05-16 10:31:15,718 - bioneuralnet.utils.preprocess - INFO - Selected 503 features by ANOVA (task=classification), 465 significant, 38 padded

Nan values in X_train_full: 0
Nan value in X_train_full after dropping: 0
X_train_full shape: (769, 2503)
Network shape: (2503, 2503)

[ ]:

from bioneuralnet.downstream_task import DPMON

save = Path("/home/vicente/Github/BioNeuralNet/TCGA_BRCA/results")
brca_pam50 = brca_pam50.rename(columns={"pam50": "phenotype"})

dpmon = DPMON(
    adjacency_matrix=A_train,
    omics_list=[meth_sel, rna_sel, mirna_sel],
    phenotype_data=brca_pam50,
    clinical_data=clinical,
    repeat_num=3,
    tune=True, gpu=True, cuda=0,
    output_dir=Path(save/"results1"),
)

predictions_df, avg_accuracy = dpmon.run()
actual = predictions_df["Actual"]
pred = predictions_df["Predicted"]
dp_acc = (accuracy_score(actual, pred), 0)
dp_f1w = (f1_score(actual, pred, average='weighted'), 0)
dp_f1m = (f1_score(actual, pred, average='macro'), 0)

print(f"DPMON results:")
print(f"Accuracy: {dp_acc[0]}")
print(f"F1 weighted: {dp_f1w[0]}")
print(f"F1 macro: {dp_f1m[0]}")

2025-05-16 15:22:49,397 - bioneuralnet.downstream_task.dpmon - INFO - Output directory set to: /home/vicente/Github/BioNeuralNet/TCGA_BRCA/results/results1
2025-05-16 15:22:49,397 - bioneuralnet.downstream_task.dpmon - INFO - Initialized DPMON with the provided parameters.
2025-05-16 15:22:49,397 - bioneuralnet.downstream_task.dpmon - INFO - Starting DPMON run.
2025-05-16 15:22:49,411 - bioneuralnet.downstream_task.dpmon - INFO - Running hyperparameter tuning for DPMON.
2025-05-16 15:22:49,412 - bioneuralnet.downstream_task.dpmon - INFO - Using GPU 0
2025-05-16 15:22:49,412 - bioneuralnet.downstream_task.dpmon - INFO - Slicing omics dataset based on network nodes.
2025-05-16 15:22:49,415 - bioneuralnet.downstream_task.dpmon - INFO - Building PyTorch Geometric Data object from adjacency matrix.
2025-05-16 15:22:49,487 - bioneuralnet.downstream_task.dpmon - INFO - Number of nodes in network: 2503
2025-05-16 15:22:49,487 - bioneuralnet.downstream_task.dpmon - INFO - Using clinical vars for node features: ['days_to_birth', 'age_at_diagnosis', 'days_to_last_followup', 'age_at_index', 'years_to_birth', 'year_of_diagnosis', 'number_of_lymph_nodes', 'date_of_initial_pathologic_diagnosis', 'histological_type_infiltrating lobular carcinoma', 'primary_diagnosis_Lobular carcinoma, NOS', 'morphology_8520/3', 'race.1_white', 'days_to_death.1', 'laterality_Right', 'primary_diagnosis_Infiltrating duct carcinoma, NOS', 'country_of_residence_at_enrollment_United States', 'sites_of_involvement_Breast, NOS', 'days_to_death', 'race_white', 'ajcc_staging_system_edition_6th']
2025-05-16 15:22:53,816 - bioneuralnet.downstream_task.dpmon - INFO - Starting hyperparameter tuning for dataset shape: (769, 2504)
2025-05-16 15:22:53,817 INFO tune.py:616 -- [output] This uses the legacy output and progress reporter, as Jupyter notebooks are not supported by the new engine, yet. For more information, please see https://github.com/ray-project/ray/issues/36949
2025-05-16 15:23:37,056 INFO tune.py:1009 -- Wrote the latest version of all result files and experiment state to '/home/vicente/ray_results/tune_dp' in 0.0110s.
2025-05-16 15:23:37,074 - bioneuralnet.downstream_task.dpmon - INFO - Best trial config: {'gnn_layer_num': 2, 'gnn_hidden_dim': 128, 'lr': 0.005072810911305633, 'weight_decay': 0.00396813275339707, 'nn_hidden_dim1': 4, 'nn_hidden_dim2': 4, 'num_epochs': 512}
2025-05-16 15:23:37,075 - bioneuralnet.downstream_task.dpmon - INFO - Best trial final loss: 1.1080904006958008
2025-05-16 15:23:37,075 - bioneuralnet.downstream_task.dpmon - INFO - Best trial final accuracy: 0.8179453836150845
2025-05-16 15:23:37,077 - bioneuralnet.downstream_task.dpmon - INFO -    gnn_layer_num  gnn_hidden_dim        lr  weight_decay  nn_hidden_dim1  \
0              2             128  0.005073      0.003968               4

   nn_hidden_dim2  num_epochs
0               4         512
2025-05-16 15:23:37,079 - bioneuralnet.downstream_task.dpmon - INFO - Best tuned parameters: {'gnn_layer_num': 2, 'gnn_hidden_dim': 128, 'lr': 0.005072810911305633, 'weight_decay': 0.00396813275339707, 'nn_hidden_dim1': 4, 'nn_hidden_dim2': 4, 'num_epochs': 512}
2025-05-16 15:23:37,079 - bioneuralnet.downstream_task.dpmon - INFO - Best tuned parameters: {'gnn_layer_num': 2, 'gnn_hidden_dim': 128, 'lr': 0.005072810911305633, 'weight_decay': 0.00396813275339707, 'nn_hidden_dim1': 4, 'nn_hidden_dim2': 4, 'num_epochs': 512}
2025-05-16 15:23:37,079 - bioneuralnet.downstream_task.dpmon - INFO - Running standard training with tuned parameters.
2025-05-16 15:23:37,079 - bioneuralnet.downstream_task.dpmon - INFO - Using GPU 0
2025-05-16 15:23:37,080 - bioneuralnet.downstream_task.dpmon - INFO - Slicing omics dataset based on network nodes.
2025-05-16 15:23:37,083 - bioneuralnet.downstream_task.dpmon - INFO - Building PyTorch Geometric Data object from adjacency matrix.
2025-05-16 15:23:37,159 - bioneuralnet.downstream_task.dpmon - INFO - Number of nodes in network: 2503
2025-05-16 15:23:37,159 - bioneuralnet.downstream_task.dpmon - INFO - Using clinical vars for node features: ['days_to_birth', 'age_at_diagnosis', 'days_to_last_followup', 'age_at_index', 'years_to_birth', 'year_of_diagnosis', 'number_of_lymph_nodes', 'date_of_initial_pathologic_diagnosis', 'histological_type_infiltrating lobular carcinoma', 'primary_diagnosis_Lobular carcinoma, NOS', 'morphology_8520/3', 'race.1_white', 'days_to_death.1', 'laterality_Right', 'primary_diagnosis_Infiltrating duct carcinoma, NOS', 'country_of_residence_at_enrollment_United States', 'sites_of_involvement_Breast, NOS', 'days_to_death', 'race_white', 'ajcc_staging_system_edition_6th']
2025-05-16 15:23:41,530 - bioneuralnet.downstream_task.dpmon - INFO - Training iteration 1/3
2025-05-16 15:23:41,563 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [1/512], Loss: 1.6182
2025-05-16 15:23:41,650 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [10/512], Loss: 1.5531
2025-05-16 15:23:41,745 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [20/512], Loss: 1.5067
2025-05-16 15:23:41,840 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [30/512], Loss: 1.4441
2025-05-16 15:23:41,939 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [40/512], Loss: 1.3666
2025-05-16 15:23:42,033 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [50/512], Loss: 1.2763
2025-05-16 15:23:42,128 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [60/512], Loss: 1.2001
2025-05-16 15:23:42,224 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [70/512], Loss: 1.1394
2025-05-16 15:23:42,317 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [80/512], Loss: 1.0970
2025-05-16 15:23:42,412 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [90/512], Loss: 1.0614
2025-05-16 15:23:42,505 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [100/512], Loss: 1.0399
2025-05-16 15:23:42,599 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [110/512], Loss: 1.0265
2025-05-16 15:23:42,693 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [120/512], Loss: 1.1995
2025-05-16 15:23:42,787 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [130/512], Loss: 1.0792
2025-05-16 15:23:42,881 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [140/512], Loss: 1.0485
2025-05-16 15:23:42,976 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [150/512], Loss: 1.0294
2025-05-16 15:23:43,077 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [160/512], Loss: 1.0155
2025-05-16 15:23:43,171 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [170/512], Loss: 1.0073
2025-05-16 15:23:43,266 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [180/512], Loss: 1.0014
2025-05-16 15:23:43,361 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [190/512], Loss: 0.9978
2025-05-16 15:23:43,455 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [200/512], Loss: 0.9956
2025-05-16 15:23:43,549 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [210/512], Loss: 0.9942
2025-05-16 15:23:43,643 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [220/512], Loss: 0.9932
2025-05-16 15:23:43,737 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [230/512], Loss: 1.0023
2025-05-16 15:23:43,831 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [240/512], Loss: 0.9951
2025-05-16 15:23:43,925 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [250/512], Loss: 0.9916
2025-05-16 15:23:44,019 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [260/512], Loss: 0.9900
2025-05-16 15:23:44,113 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [270/512], Loss: 0.9888
2025-05-16 15:23:44,207 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [280/512], Loss: 0.9878
2025-05-16 15:23:44,302 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [290/512], Loss: 0.9872
2025-05-16 15:23:44,396 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [300/512], Loss: 1.0332
2025-05-16 15:23:44,490 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [310/512], Loss: 1.0558
2025-05-16 15:23:44,584 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [320/512], Loss: 1.0264
2025-05-16 15:23:44,678 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [330/512], Loss: 1.0068
2025-05-16 15:23:44,772 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [340/512], Loss: 0.9968
2025-05-16 15:23:44,866 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [350/512], Loss: 0.9917
2025-05-16 15:23:44,963 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [360/512], Loss: 0.9890
2025-05-16 15:23:45,059 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [370/512], Loss: 1.0214
2025-05-16 15:23:45,153 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [380/512], Loss: 1.0091
2025-05-16 15:23:45,247 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [390/512], Loss: 0.9997
2025-05-16 15:23:45,341 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [400/512], Loss: 0.9948
2025-05-16 15:23:45,436 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [410/512], Loss: 0.9907
2025-05-16 15:23:45,530 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [420/512], Loss: 0.9878
2025-05-16 15:23:45,624 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [430/512], Loss: 0.9851
2025-05-16 15:23:45,717 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [440/512], Loss: 0.9828
2025-05-16 15:23:45,815 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [450/512], Loss: 0.9802
2025-05-16 15:23:45,909 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [460/512], Loss: 0.9773
2025-05-16 15:23:46,003 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [470/512], Loss: 0.9736
2025-05-16 15:23:46,097 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [480/512], Loss: 0.9691
2025-05-16 15:23:46,192 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [490/512], Loss: 0.9645
2025-05-16 15:23:46,287 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [500/512], Loss: 0.9643
2025-05-16 15:23:46,384 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [510/512], Loss: 1.0325
2025-05-16 15:23:46,407 - bioneuralnet.downstream_task.dpmon - INFO - Training Accuracy: 0.6658
2025-05-16 15:23:46,409 - bioneuralnet.downstream_task.dpmon - INFO - Model saved to /home/vicente/Github/BioNeuralNet/TCGA_BRCA/results/results1/dpm_model_iter_1.pth
2025-05-16 15:23:46,413 - bioneuralnet.downstream_task.dpmon - INFO - Training iteration 2/3
2025-05-16 15:23:46,435 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [1/512], Loss: 1.6234
2025-05-16 15:23:46,531 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [10/512], Loss: 1.5598
2025-05-16 15:23:46,626 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [20/512], Loss: 1.5203
2025-05-16 15:23:46,720 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [30/512], Loss: 1.4727
2025-05-16 15:23:46,814 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [40/512], Loss: 1.4078
2025-05-16 15:23:46,908 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [50/512], Loss: 1.3256
2025-05-16 15:23:47,002 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [60/512], Loss: 1.2342
2025-05-16 15:23:47,096 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [70/512], Loss: 1.1537
2025-05-16 15:23:47,190 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [80/512], Loss: 1.1014
2025-05-16 15:23:47,284 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [90/512], Loss: 1.0631
2025-05-16 15:23:47,378 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [100/512], Loss: 1.0422
2025-05-16 15:23:47,472 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [110/512], Loss: 1.0257
2025-05-16 15:23:47,565 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [120/512], Loss: 1.0252
2025-05-16 15:23:47,659 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [130/512], Loss: 1.0136
2025-05-16 15:23:47,757 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [140/512], Loss: 1.0068
2025-05-16 15:23:47,852 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [150/512], Loss: 1.0016
2025-05-16 15:23:47,946 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [160/512], Loss: 0.9965
2025-05-16 15:23:48,040 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [170/512], Loss: 1.0520
2025-05-16 15:23:48,134 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [180/512], Loss: 1.0251
2025-05-16 15:23:48,229 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [190/512], Loss: 1.0158
2025-05-16 15:23:48,322 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [200/512], Loss: 1.0110
2025-05-16 15:23:48,416 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [210/512], Loss: 1.0074
2025-05-16 15:23:48,523 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [220/512], Loss: 1.0034
2025-05-16 15:23:48,618 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [230/512], Loss: 1.0010
2025-05-16 15:23:48,712 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [240/512], Loss: 0.9988
2025-05-16 15:23:48,806 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [250/512], Loss: 0.9974
2025-05-16 15:23:48,900 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [260/512], Loss: 0.9960
2025-05-16 15:23:48,994 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [270/512], Loss: 0.9946
2025-05-16 15:23:49,089 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [280/512], Loss: 0.9932
2025-05-16 15:23:49,183 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [290/512], Loss: 0.9924
2025-05-16 15:23:49,277 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [300/512], Loss: 0.9905
2025-05-16 15:23:49,373 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [310/512], Loss: 0.9880
2025-05-16 15:23:49,467 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [320/512], Loss: 0.9850
2025-05-16 15:23:49,561 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [330/512], Loss: 0.9789
2025-05-16 15:23:49,656 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [340/512], Loss: 0.9995
2025-05-16 15:23:49,750 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [350/512], Loss: 0.9807
2025-05-16 15:23:49,844 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [360/512], Loss: 0.9594
2025-05-16 15:23:49,938 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [370/512], Loss: 0.9547
2025-05-16 15:23:50,031 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [380/512], Loss: 0.9526
2025-05-16 15:23:50,126 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [390/512], Loss: 0.9514
2025-05-16 15:23:50,220 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [400/512], Loss: 0.9506
2025-05-16 15:23:50,314 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [410/512], Loss: 0.9505
2025-05-16 15:23:50,417 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [420/512], Loss: 0.9510
2025-05-16 15:23:50,511 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [430/512], Loss: 0.9494
2025-05-16 15:23:50,605 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [440/512], Loss: 0.9485
2025-05-16 15:23:50,699 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [450/512], Loss: 0.9477
2025-05-16 15:23:50,793 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [460/512], Loss: 0.9473
2025-05-16 15:23:50,887 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [470/512], Loss: 0.9470
2025-05-16 15:23:50,981 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [480/512], Loss: 0.9455
2025-05-16 15:23:51,074 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [490/512], Loss: 0.9464
2025-05-16 15:23:51,168 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [500/512], Loss: 0.9468
2025-05-16 15:23:51,262 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [510/512], Loss: 0.9451
2025-05-16 15:23:51,285 - bioneuralnet.downstream_task.dpmon - INFO - Training Accuracy: 0.9545
2025-05-16 15:23:51,287 - bioneuralnet.downstream_task.dpmon - INFO - Model saved to /home/vicente/Github/BioNeuralNet/TCGA_BRCA/results/results1/dpm_model_iter_2.pth
2025-05-16 15:23:51,291 - bioneuralnet.downstream_task.dpmon - INFO - Training iteration 3/3
2025-05-16 15:23:51,309 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [1/512], Loss: 1.6141
2025-05-16 15:23:51,394 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [10/512], Loss: 1.5460
2025-05-16 15:23:51,488 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [20/512], Loss: 1.5051
2025-05-16 15:23:51,582 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [30/512], Loss: 1.4469
2025-05-16 15:23:51,676 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [40/512], Loss: 1.3636
2025-05-16 15:23:51,771 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [50/512], Loss: 1.2702
2025-05-16 15:23:51,865 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [60/512], Loss: 1.1908
2025-05-16 15:23:51,963 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [70/512], Loss: 1.1343
2025-05-16 15:23:52,057 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [80/512], Loss: 1.1024
2025-05-16 15:23:52,152 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [90/512], Loss: 1.0812
2025-05-16 15:23:52,247 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [100/512], Loss: 1.1784
2025-05-16 15:23:52,343 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [110/512], Loss: 1.1234
2025-05-16 15:23:52,437 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [120/512], Loss: 1.1016
2025-05-16 15:23:52,531 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [130/512], Loss: 1.0903
2025-05-16 15:23:52,625 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [140/512], Loss: 1.0835
2025-05-16 15:23:52,724 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [150/512], Loss: 1.0627
2025-05-16 15:23:52,818 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [160/512], Loss: 1.0492
2025-05-16 15:23:52,911 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [170/512], Loss: 1.0344
2025-05-16 15:23:53,005 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [180/512], Loss: 1.0256
2025-05-16 15:23:53,099 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [190/512], Loss: 1.0206
2025-05-16 15:23:53,193 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [200/512], Loss: 1.0164
2025-05-16 15:23:53,287 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [210/512], Loss: 1.0128
2025-05-16 15:23:53,383 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [220/512], Loss: 1.0165
2025-05-16 15:23:53,480 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [230/512], Loss: 1.0111
2025-05-16 15:23:53,574 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [240/512], Loss: 1.0075
2025-05-16 15:23:53,670 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [250/512], Loss: 1.0027
2025-05-16 15:23:53,764 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [260/512], Loss: 0.9996
2025-05-16 15:23:53,868 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [270/512], Loss: 0.9979
2025-05-16 15:23:53,962 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [280/512], Loss: 1.0081
2025-05-16 15:23:54,056 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [290/512], Loss: 1.0096
2025-05-16 15:23:54,150 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [300/512], Loss: 1.0006
2025-05-16 15:23:54,243 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [310/512], Loss: 0.9946
2025-05-16 15:23:54,337 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [320/512], Loss: 0.9909
2025-05-16 15:23:54,431 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [330/512], Loss: 0.9882
2025-05-16 15:23:54,525 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [340/512], Loss: 0.9877
2025-05-16 15:23:54,619 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [350/512], Loss: 0.9849
2025-05-16 15:23:54,712 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [360/512], Loss: 0.9833
2025-05-16 15:23:54,806 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [370/512], Loss: 0.9818
2025-05-16 15:23:54,900 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [380/512], Loss: 0.9807
2025-05-16 15:23:54,993 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [390/512], Loss: 1.0491
2025-05-16 15:23:55,087 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [400/512], Loss: 0.9906
2025-05-16 15:23:55,181 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [410/512], Loss: 0.9831
2025-05-16 15:23:55,275 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [420/512], Loss: 0.9797
2025-05-16 15:23:55,370 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [430/512], Loss: 0.9777
2025-05-16 15:23:55,464 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [440/512], Loss: 0.9759
2025-05-16 15:23:55,558 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [450/512], Loss: 0.9749
2025-05-16 15:23:55,652 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [460/512], Loss: 0.9742
2025-05-16 15:23:55,747 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [470/512], Loss: 0.9737
2025-05-16 15:23:55,851 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [480/512], Loss: 0.9738
2025-05-16 15:23:55,944 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [490/512], Loss: 0.9735
2025-05-16 15:23:56,038 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [500/512], Loss: 0.9727
2025-05-16 15:23:56,132 - bioneuralnet.downstream_task.dpmon - INFO - Epoch [510/512], Loss: 0.9720
2025-05-16 15:23:56,155 - bioneuralnet.downstream_task.dpmon - INFO - Training Accuracy: 0.9558
2025-05-16 15:23:56,157 - bioneuralnet.downstream_task.dpmon - INFO - Model saved to /home/vicente/Github/BioNeuralNet/TCGA_BRCA/results/results1/dpm_model_iter_3.pth
2025-05-16 15:23:56,161 - bioneuralnet.downstream_task.dpmon - INFO - Best Accuracy: 0.9558
2025-05-16 15:23:56,162 - bioneuralnet.downstream_task.dpmon - INFO - Average Accuracy: 0.8587
2025-05-16 15:23:56,162 - bioneuralnet.downstream_task.dpmon - INFO - Standard Deviation of Accuracy: 0.1670

DPMON results:
Accuracy: 0.9557867360208062
F1 weighted: 0.9360974742812752
F1 macro: 0.7772360237077294