Minor technical notes on TFBS models presented in HOCOMOCO-v10 1. All the matrices were produced by ChIPMunk motif discovery tool. It accounts for prior "quality" information for each sequence in the input set. The resulting words in the alignment are weighted and thus the position count matrix (PCM) is weighted PCM and may contain floating point numbers. These numbers are treated as counts in a PCM-to-PWM conversion scheme. In case integer counts are required, we suggest using HOCOMOCO downloads in TRANSFAC format (see Downloads) where the weighted counts were carefully converted to the nearest integers maintaining consistency of the resulting PCMs. 2. The PCM to PWM conversion used in HOCOMOCO-v10 is the same as in v9, the exact formula can be found, e.g., in MACRO-APE manual, page 20-21 in http://opera.autosome.org/downloads/MACRO-PERFECTOS-APE_manual.pdf 3. The plain files with matrices always use alphabetical order of nucleotides (ACGT) and dinucleotides (AA,AC,AG...TT). 4. The TFBS alignments (*.words) provide information on particular sequence segments used by ChIPMunk to derive the positional matrices. Many models were inherited from HOCOMOCO v9 where some binding sites, in particular, were mined from the proprietary TRANSFAC database. The respective aligned binding sites were masked with 'X's.