Find the shortcut. Learn the biology.
Automated quality control for genomic machine learning datasets: scores the biases, duplicates and data leakage a classifier could exploit before you train on it.
Searching test set enhancer_promoter_interaction_K562_test.csv.gz against train set enhancer_promoter_interaction_K562_train.csv.gz
Data Leakage:!WarningHow a flag is decided →
gb-qc v1.0.0 · run 2026-09-15 22:31:15

Basic Descriptive Statistics

Filename enhancer_promoter_interaction_K562_test.csv.gz enhancer_promoter_interaction_K562_train.csv.gz
Number of sequences 2000 10000
Minimum length 5000 5000
Mean length 5000.00 5000.00
Maximum length 5000 5000

Every test sequence searched against the train set with MMseqs2, and scored on its best hit. Similarity is min(query coverage, target coverage) × percent identity: how much of the shorter sequence the alignment spans, times how much of that matches. The histogram is the distribution of those best hits.

The flag is the percentage of test sequences at or above the threshold: Pass at 0%, Warning below 2%, Fail at 2% or more. A test sequence that near-duplicates a training one is one the model has already seen, so the accuracy it earns there is not evidence. What to do about it →

The panel under the histogram lists those alignments, up to the first 100; every one of them, listed or not, is exported to mmseqs/mmseqs2_search_result.tsv beside this report. Its columns are:

Query (Q)
Identifier of the test sequence: seq_<i>_test, where i is its 0-based position in the input test file.
Target (T)
Identifier of the train sequence: seq_<i>_train, where i is its 0-based position in the input train file.
Q Cov.
Fraction of the query sequence covered by the alignment (0–1).
T Cov.
Fraction of the target sequence covered by the alignment (0–1).
% Identity
Percent identical aligned positions in the aligned region.
% Similarity
Similarity score used for leakage detection, calculated as min(Q Cov., T Cov.) × % Identity.
E-value
MMseqs2 E-value (lower is more "significant").
Alignment
Click “Show” to expand the alignment visualisation for a row.
Filename enhancer_promoter_interaction_K562_test.csv.gz enhancer_promoter_interaction_K562_train.csv.gz
Data Leakage (percentage) 1.25% 0.26%
Data Leakage (count) 25 26
Similarity Histogram
26 high-similarity alignments
Query (Q) Target (T) Q Cov. T Cov. % Identity % Similarity E-value Alignment
seq_612_test seq_4394_train 0.99 0.99 99.5 98.50 0.00e+00
seq_823_test seq_3249_train 0.99 0.99 99.5 98.50 0.00e+00
seq_802_test seq_5271_train 0.99 0.99 99.5 98.50 0.00e+00
seq_1715_test seq_8986_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1873_test seq_735_train 0.99 0.99 99.4 98.41 0.00e+00
seq_166_test seq_5633_train 0.99 0.99 99.4 98.41 0.00e+00
seq_239_test seq_7925_train 0.99 0.99 99.4 98.41 0.00e+00
seq_8_test seq_9939_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1837_test seq_1997_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1298_test seq_7697_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1185_test seq_7061_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1712_test seq_9856_train 0.99 0.99 99.4 98.41 0.00e+00
seq_190_test seq_9517_train 0.99 0.99 99.4 98.41 0.00e+00
seq_706_test seq_8495_train 0.99 0.99 99.4 98.41 0.00e+00
seq_633_test seq_3711_train 0.99 0.99 99.4 98.41 0.00e+00
seq_665_test seq_1747_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1725_test seq_2599_train 0.99 0.99 99.4 98.41 0.00e+00
seq_675_test seq_9225_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1417_test seq_2422_train 0.99 0.99 99.4 98.41 0.00e+00
seq_1576_test seq_1102_train 0.99 0.99 99.3 98.31 0.00e+00
seq_1553_test seq_2519_train 0.99 0.99 99.3 98.31 0.00e+00
seq_1843_test seq_896_train 0.99 0.99 99.2 98.21 0.00e+00
seq_1283_test seq_7591_train 0.99 0.99 99.2 98.21 0.00e+00
seq_227_test seq_9858_train 0.99 0.99 99.1 98.11 0.00e+00
seq_132_test seq_9397_train 0.98 0.98 99.0 97.02 0.00e+00
seq_227_test seq_4414_train 0.96 0.96 98.6 94.66 0.00e+00