This occurred when the Bloom filter yielded false positives.
The k-mer academy contained a k-mer with a coverage of 2 (should be 1, but it is a false positive)
Then, in the de Bruijn graph, the real coverage was 1.
Thus the error.
The academy was removed in this merge: 290544f2628ffad344246dd36e6fdee595976f50
This occurred when the Bloom filter yielded false positives. The k-mer academy contained a k-mer with a coverage of 2 (should be 1, but it is a false positive)
Then, in the de Bruijn graph, the real coverage was 1.
Thus the error.
The academy was removed in this merge: 290544f2628ffad344246dd36e6fdee595976f50