2 2 Background: 2 3 Methods for predicting nucleosome positioning include bioinformatics, 2 4 biophysical, and combined approaches. An advantage of bioinformatics 2 5 methods, which are based on in vivo nucleosome maps, is the use of natural 2 6 sequences that may contain previously unknown elements involved in 2 7 nucleosome positioning in vivo. The accuracy of such prediction attempts 2 8 reflects the genomic coordinate resolution of the nucleosome maps applied. 2 9 Thus, nucleosome maps constructed using Micrococcus nuclease (MNase), 3 0 which has a strong preference for A/T-rich sequences, may not be appropriate 3 1 for this purpose. In addition to MNase seq-based maps, base pair-resolution 3 2 chemical maps of in vivo nucleosomes from three different species (budding and 3 3 2 fission yeasts, and mouse) are currently available. However, these maps have 3 4 yet to be integrated into publicly available computational methods. 3 5 3 6Results:3 7We developed a Bioconductor package (named nuCpos) that uses chemical 3 8 maps to train duration hidden Markov models (dHMMs) to predict nucleosome 3 9positioning. The accuracy of chemical map-based prediction was higher than 4 0 that of the previously developed MNase seq-based approach. With our method, 4 1 predicted nucleosome occupancy reasonably matched in vivo observations and 4 2 was not affected by A/T nucleotide frequency. Effects of genetic alterations on 4 3 nucleosome positioning that had been observed in living yeast could also be 4 4 predicted. In addition to dHMM-based prediction, nuCpos can calculate 4 5 individual histone binding affinity (HBA) scores for given 147-bp sequences to 4 6 examine their suitability for nucleosome formation. Local HBA scores for 13 4 7 overlapping nucleosomal DNA subsegments can also be calculated. HBA and 4 8 local HBA scores for various sequences agreed well with previous in vitro and in 4 9vivo studies. Furthermore, our results suggest that nucleosomal subsegments 5 0 that are disfavored in different rotational settings contribute to the defined 5 1 positioning of nucleosomes. 5 2 5 3 Conclusions: 5 4Our results demonstrate that chemical map-based statistical models are 5 5 beneficial for studying nucleosomal DNA features. Studies employing nuCpos 5 6software can enhance understanding of chromatin regulation and the 5 7