Summary

Basic Descriptive Statistics
! Data Leakage
Documentation GitHub
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 H3K9me3_test.csv.gz against train set H3K9me3_train.csv.gz
Data Leakage:!WarningHow a flag is decided →
gb-qc v1.0.0 · run 2026-09-15 22:31:20

Basic Descriptive Statistics

Filename H3K9me3_test.csv.gz H3K9me3_train.csv.gz
Number of sequences 850 27438
Minimum length 1000 1000
Mean length 1000.00 1000.00
Maximum length 1000 1000
!

Data Leakage

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 H3K9me3_test.csv.gz H3K9me3_train.csv.gz
Data Leakage (percentage) 1.06% 0.04%
Data Leakage (count) 9 11
Similarity Histogram
15 high-similarity alignments
Query (Q) Target (T) Q Cov. T Cov. % Identity % Similarity E-value Alignment
seq_830_test seq_21993_train 1.00 1.00 97.9 97.61 0.00e+00
seq_652_test seq_27281_train 0.99 0.99 97.9 97.31 0.00e+00
seq_607_test seq_14956_train 0.99 0.99 96.9 96.42 0.00e+00
seq_505_test seq_14510_train 0.99 0.99 97.1 96.23 0.00e+00
seq_533_test seq_14956_train 0.99 0.99 97.5 96.14 0.00e+00
seq_847_test seq_25667_train 0.99 0.99 96.5 95.73 0.00e+00
seq_533_test seq_18514_train 0.96 0.96 96.8 93.12 0.00e+00
seq_533_test seq_19309_train 0.95 0.95 97.3 92.92 0.00e+00
seq_652_test seq_18982_train 0.97 0.97 94.8 92.43 0.00e+00
seq_607_test seq_18514_train 0.95 0.95 96.4 91.87 0.00e+00
seq_778_test seq_14510_train 0.95 0.95 96.2 91.58 0.00e+00
seq_607_test seq_19309_train 0.95 0.95 96.5 91.29 0.00e+00
seq_800_test seq_15613_train 0.97 0.97 94.0 90.99 0.00e+00
seq_677_test seq_15811_train 0.99 0.99 91.8 90.79 0.00e+00
seq_677_test seq_20450_train 1.00 0.99 91.4 90.49 0.00e+00