Features which cover an entire circular sequence are not correctly shifted

[amijalis]

I noticed today that a feature spanning an entire circular sequence is not properly shifted with Dseqrecord.shifted().

Reproduce with:

test_seq = 'ATCGATCGATCGATCGATCGATCGATCGATCGATCG'

# length of test_seq is 36.

test_seq_dseqrecord = Dseqrecord(test_seq).looped()
test_seq_dseqrecord.add_feature(0,36,type_='test', label='test')

print(f'Features before shifting: {test_seq_dseqrecord.features}')

test_seq_shifted = test_seq_dseqrecord.shifted(10)

print(f'Features after shifting: {test_seq_shifted.features}')

Output:

Features before shifting: [SeqFeature(SimpleLocation(ExactPosition(0), ExactPosition(36), strand=1), type='test', qualifiers=...)]

Features after shifting: []

The expected behavior is observed if the feature is annotated on the interval from [1: len(seq)]:

from pydna.dseqrecord import Dseqrecord
test_seq = 'ATCGATCGATCGATCGATCGATCGATCGATCGATCG'

# length of test_seq is 36.

test_seq_dseqrecord = Dseqrecord(test_seq).looped()
test_seq_dseqrecord.add_feature(1,36,type_='test', label='test')

print(f'Features before shifting: {test_seq_dseqrecord.features}')

test_seq_shifted = test_seq_dseqrecord.shifted(10)

print(f'Features after shifting: {test_seq_shifted.features}')

Output:

Features before shifting: [SeqFeature(SimpleLocation(ExactPosition(1), ExactPosition(36), strand=1), type='test', qualifiers=...)]

Features after shifting: [SeqFeature(CompoundLocation([SimpleLocation(ExactPosition(27), ExactPosition(36), strand=1), SimpleLocation(ExactPosition(0), ExactPosition(26), strand=1)], 'join'), type='test', location_operator='join', qualifiers=...)]

The issue seems to be in dseqrecord.py where the newstart and newend variables are computed. If the feature length is equal to the sequence length, newstart and newend end up being the same, and the feature isn't added.

pydna/src/pydna/dseqrecord.py

Lines 1092 to 1094 in f8d4850

	for location in shiftedparts:
	newstart = location.start % ln
	newend = location.end % ln

[BjornFJohansson]

Thank you for your input. There are a couple of open bugs around features at the moment. Ill look into these next week.

[BjornFJohansson]

I think I have this solved. look out for a new release.

[amijalis]

Thank you! looking forward to it!

[BjornFJohansson]

Solved in 5.2.0 I hope. Feel free to reopen if not.