Correction to Nature Microbiology https://doi.org/10.1038/s41564-018-0162-2, published 21 May 2018.

In the version of this Article originally published, the reference genome assembly for Plasmodium billcollinsi included an error that affected four small regions. The authors’ workflow for producing assemblies included a correction step in which reads from the original sample (containing one or more species) were aligned to the assembly. In this particular assembly—originating from a mixed Plasmodium gaboni/P. billcollinsi infection—reads from P. gaboni were more abundant than P. billcollinsi, but at most sites could be filtered by mapping to the P. gaboni co-infection. For the sites in question, there were gaps in the P. gaboni assembly and the more numerous P. gaboni reads were therefore erroneously used to morph the P. billcollinsi reference at these points. In subsequent analyses, the authors misinterpreted these four instances as introgression from the P. gaboni lineage into P. billcollinsi. Other analyses involving the four genes present in these regions (PBILG01_0522300, PBILG01_0613600, PBILG01_1327900 and PBILG01_1328500) should also be regarded as incorrect for P. billcollinsi. The P. billcollinsi reference assembly has now been corrected and can be accessed using the NCBI accession number GCA_900257145.2. The authors have competitively mapped the original reads against P. gaboni and P.billcollinsi and have found no further regions where P. gaboni preferentially map to the P. billcollinsi reference.

Furthermore, in the ‘P. falciparum-specific evolution’ section of the main text, the sentence ‘Two genes (rop14 and PF3D7_0609900) were significant in both tests.’ has been removed, as the results in Supplementary Table 4 represent the authors’ raw analysis and were not corrected for multiple comparisons. The word ‘significant’ has also been removed from the sentence beginning ‘Other genes, expressed in...’ in the same paragraph.

The authors thank L. Plenderleith and P. Sharp for their comments and for identifying the misassembly that allowed the authors to correct some errors in their interpretation of the data.