When you use the Elongation method of condensation for Illumina data, SOLiD System data, or Ion Torrent data, overlapping reads are not merged. Instead, a new elongated read with errors corrected is created for each read in the subgroup. Because a given read is likely to match more than one anchor sequence, all instances of a given read are pooled “as is” into multiple subgroups. These corrected and elongated reads are then compared to each other to produce a single consensus sequence. Reads that do not match any of the indices are not removed as in consolidation, but instead, are kept in the output file.