Chemokine genes typically consist of 3�C4 exons, which coincides with what we observe for the B42 and N73 genes. The results obtained for the discovery of B42 and N73 as putative new human chemokines and their further characterization are described in detail in the following sections. B42 is 81 amino acids long and contains 6 cysteine residues. By using standard sequence-based methods, we found no significant sequence homology of B42 to any previously characterized protein. B42 was submitted to UniProt/TrEMBL together with four zinc finger proteins. The B42 gene was first automatically labeled as zinc finger protein ZNF528 in Ensembl because its protein coding region was located within the untranslated region of the Kruppel-like zinc finger gene ZNF528. The translated B42 protein does not share sequence LDN-193189 identity to any characterized zinc finger protein, however, many zinc finger genes are located in close proximity to the B42 gene on the chromosome. Since Ensembl release 55, it was renamed to ZNF578, which, according to the Ensembl annotation, does not code for a protein, as it is labeled as ��processed transcript��. The gene annotation of the human B42 gene produced by the automatic Ensembl genebuild prediction method did not detect the B42 gene in more recent Ensembl releases for reasons not specified on their website, although the genomic sequence is identical to the sequence of previous versions. Interestingly, the orthologous chimpanzee and orangutan genes of B42 are still annotated as novel protein coding genes in the current Ensembl release 64. In order to quantify the residue energetic contributions in the B42 chemokine-like model, we calculated contact energies per residue and compared them to those of the vMIP-I model. The contact energy decomposition plot for B42 compared to vMIP-I shows a general agreement in the energetic contributions of the corresponding residues. In particular, it can be observed that the secondary structure regions and the core of the protein present a good agreement of the contact energies, indicating that the LY294002 in vivo packing of the core of B42 is comparable to the packing of the vMIP-I template. We calculated the coefficient of determination from the contact energy pairs of the B42 model and the reference model of vMIP-I, and we obtained an R2 value of 0.63. We also did this analysis for the vMIP-I using as reference a mutant of CXCL8, and we obtained an R2 value of 0.43. We selected this CXCL8 mutant as reference, because it has one disulfide bond at a non-canonical position when compared to the rest of the chemokine protein family, and it has low sequence identity to vMIP-I. The R2 value obtained for B42 is higher, which is indicative of the good packing of the core in the B42 model compared to known chemokines and gives high confidence to its predicted structural resemblance towards the IL8-like chemokine fold.