Additional file 1 of Genomic data integration and user-defined sample-set extraction for population variant analysis

Additional file 1. Example of translation from VCF into GDM format for genomic region data: This .xlsx (MS Excel) spreadsheet exemplifies the transformation of the original 1KGP mutations—expressed in VCF format—into GDM genomic regions. As a demonstrative example, some variants about chromosome X have been selected from the source data (in VCF format) and listed in the first table at the top of the file. The values of columns #CHROM, POS, REF and ALT appear as in the source. We removed the details that are unnecessary for the transformation from the column INFO. In the column FORMAT it is indicated exclusively the value “GT”, meaning that the next columns contain only the genotype of the samples (this and other conventions are expressed in the VCF specification document and in the header section of each VCF file). In multiallelic variants (examples e, f.1 and f.2), the genotype indicates with a number which of the alternative alleles in ALT is present in the corresponding samples (e.g., the number 2 means that the second variant is present); otherwise, it only assumes the values 0—mutation absent, or 1—the mutation is present. Additionally, the genotype indicates whether one or both chromosome copies contain the mutation and which one, i.e., the left one or the right one; the mutated alleles are normally separated by a pipe (“|”), if not otherwise specified in the header section; we do not know which chromosome copy is maternal or paternal, but as the 1KGP mutations are “phased”, we know that the “left chromosome” is the same in every mutation located in the same chromosome of the same donor. As in this example we have only one column after the FORMAT one, the mutations described are relative to only one sample, called “HG123456”. Actually, this sample does not exist in the source, but serves the purpose of demonstrating several mutation types that are found in the original data. The table reports six variants in VCF format, with the last one repeated two times to show how different values of genotype lead to a different translation (indeed, examples f.1 and f.2 differ only for the last column). Below in the same file, the same variants appear converted in GDM format. The transformation outputs the chr, left, right, strand, AL1, AL2, ref, alt, mut_type and length columns. The value of strand is positive in every mutation, as clarified by the 1KGP Consortium after the release of the data collections. Values of AL1 and AL2 express on which chromatid the mutation occur and depend on the value of the original genotype (column HG123456). The values of the other columns, namely chr, left, right, ref, alt, mut_type and length, are obtained from the variant original values after the split of multi-allelic variants, the transformation of the original position into 0-based coordinates, and the removal of repeated nucleotide bases from the original REF and ALT columns. In 0-based coordinates, a nucleotide base occupies the space between the coordinates x and x + 1. So, SNPs (examples a and f.2) are encoded as the replacement of ref at position between left and right with alt. Insertions (examples c and f.1) are described as the addition of the sequence of bases in alt at the position indicated in left and right, i.e., in between two nucleotide bases. Deletions (example b) are represented as the substitution of ref between positions left and right with an empty value (alt is indeed empty in this case). Finally, structural variants (examples d and e) such as copy number variations and large deletions have an empty ref because, according to the VCF specification document, the original column REF reports a nucleotide (called padding-base) that is located before the scope of the variant on the genome and is unnecessary in a 0-based representation. In this file, we reported only the columns relevant for the understanding of the transformation method regarding the mutation coordinates, reference and alternative alleles. Actually, in addition to the ones reported in the second table, the transformation adds some more columns, called as the attributes in the original INFO column to capture a selection of the attributes present in the original file.