Fix reverse complement calculation in CDSSequence.getCodingSequence (#1016) - #1160

Fixes #1016

Problem

When CDSSequence.getCodingSequence() is called on a negative-strand transcript/gene, it previously resulted in a sequence that was only reversed rather than reverse-complemented.

This occurred because this.getSequenceAsString(getBioBegin(), getBioEnd(), getStrand()) delegates to SimpleLocation.getSubSequence(), which invokes AbstractLocation.reverseSequence(). For Strand.NEGATIVE, AbstractLocation.reverseSequence() already wraps the sequence in both ReversedSequenceView and ComplementSequenceView, successfully generating the reverse complement.

However, CDSSequence.getCodingSequence() then executed an additional if (getStrand() == Strand.NEGATIVE) block that complemented every nucleotide a second time. Because $\text{complement}(\text{complement}(N)) = N$, this second complementation inverted the nucleotides back to their original bases while leaving the sequence reversed (e.g. producing CCCCCAAAAA instead of GGGGGTTTTT).

Solution

• Removed the redundant second complementation block in CDSSequence.getCodingSequence().
• Removed now-unused imports (CompoundSet, NucleotideCompound) in CDSSequence.java.
• Enabled the existing disabled unit test TranscriptSequenceTest.getCDNASeqNegativeStrand() and asserted both CDSSequence.getCodingSequence() and TranscriptSequence.getDNACodingSequence() return the proper reverse complement.
• Verified all 536 tests in biojava-core pass with zero regressions.