Cyanophages (cyanobacterial viruses) are important agents of horizontal gene transfer among marine cyanobacteria, the numerically dominant photosynthetic organisms in the oceans.
http://plosbiology.org/article/info:doi/10.1371/journal.pbio.0040234#article1.front1.article-meta1.abstract1.p1

Recent whole genome cyanophage sequencing (Sullivan et al 2010) has revealed that the sequences of some genes do not match those published previously as single gene sequences derived from DNA fragments PCR-amplified from phage isolates. Specifically, four psbA gene sequences reported previously in Sullivan & Lindell et al. (2006) were incorrect: S-SSM5 (gene GI:95115381, protein GI:95115382), P-SSP5 (gene GI:95115369, protein GI:95115370; this 'P-SSP5' sequence was only found in the phage P-SSP3 genome), P-SSP6 (gene GI:95115367, protein GI:95115368), and P-SSP3 (gene GI:95115373, protein GI:95115374; this ‘P-SSP3’ sequence was not found in any genome sequenced). The Genbank entries have now been corrected. The revised sequences do not alter the overall psbA tree topology, intragenic recombination findings, or the conclusions of the paper.

The source of error for these gene sequences is not certain, but could have its origin in either a clerical error, PCR contamination, and/or mixed phage stocks. On this latter point, it has recently come to light that triplicate plaque purifications do not always result in pure stocks for these cyanophages, possibly due to the fluid consistency of the agarose plates (0.28% agarose) required for growth of the host. More reliable purification procedures have subsequently been developed. However, as mixed phage stock preparations could influence host range, we recommend that published host range data (Sullivan et al 2003) be verified before their use in further research.