The package contains 8 BAM files, 1 per sequencing run. Each BAM file was obtained by (1) aligning the reads (paired-end) to the full hg19 genome with TopHat2, and then (2) subsetting to keep only alignments on chr14. See accession number E-MTAB-1147 in the ArrayExpress database for details about the experiment, including links to the published study (by Zarnack et al., 2012) and to the FASTQ files.
providing 
.