Breast cancer is one of the major causes of cancer-related deaths worldwide and the most common cancer among women . Metastatis to distant organs and lymph nodes represents a major problem, usually leading to high mortality. The investigation of breast cancer-associated genes for early detection or therapeutic targeting could potentially improve the survival rates of breast cancer patients.
MicroRNAs (miRNAs) are small RNA molecules with important regulatory functions in several physiological activities . MicroRNAs are processed from primary transcripts (pri-miRNAs) in 2 maturation steps. First, the pri-miRNAs are processed by Drosha, forming the precursor miRNAs (pre-miRNAs), composed of a 5p arm, a 3p arm, and a terminal loop, approximately 70 nucleotides in length. Following the transport of pre-miRNAs to the cytoplasm by exportin 5, they are further processed by Dicer to release the terminal loop and the duplex (5p arm/3p arm), 22 nucleotides in length. The 5p arm/3p arm of the duplex is unwound at the end because of weaker hydrogen binding. The 5p or the 3p arm is selectively loaded into the RNA-induced silencing complex (RISC) and serves as mature miRNA [3–5]. Recent studies described a phenomenon in which RNA editing or nucleotide addition generated 3' end sequence variants of miRNAs [6–13]. Fernandez-Valverde et al.  reported that miR-282 and miR-312a are enriched for 3' adenosine additions during early embryonic development, which increases miRNA stability or enhances miRNA and mRNA interaction.
MicroRNAs exert their effects by repressing their target genes. They downregulate target gene expression by repressing translation or by degrading mRNAs. Previous studies reported that miRNAs play important roles in the oncogenesis pathway [14–18]. The tumor-associated miRNAs were either tumor repressors, preferentially expressed in normal tissue, or onco-miRNAs, preferentially expressed in tumor tissue. These are aberrantly expressed in human breast cancer, including miR-9, miR-21, miR-31, miR-34a, miR-155, miR-200, miR-205, miR-206, and miR-335 [19–22]. Although several studies have investigated the functions of miRNAs in breast tumors, these only included a small fraction of existing miRNAs [23–26]. Using miRNA profiling approach, numerous breast cancer-associated miRNAs were identified [19, 27–30]. Ryu et al.  identified 189 candidate novel microRNAs in human breast cancer cell lines by deep sequencing technology. Therefore, emerging NGS technologies can be used not only to identify novel miRNAs, but can also be applied in several miRNA-associated studies.
In the studies using NGS data for miRNA profiling, it is usually observed that miRNA sequence reads exist as isoforms, named isomiRs, with position and length shift compared with the reference miRNAs . Recently, more and more studies worked on the isomiR issues, such as isomiR pattern preferences in specific libraries, target gene selection difference between different isomiRs and so on [6, 32]. Therefore, NGS data provides a good resource for miRNA expression profiling and isomiR related studies. In 2011, Farazi et al.,  used NGS data to determine miRNA expression profiles in breast tissues with differing tumor malignancies. They focused on the relevance of specific miRNAs and the tumor malignancy type, without providing further experimental validation. The present study applied their NGS data to conduct analysis of miRNA-associated changes in breast cancer, including differential miRNA expression, position shifts in isomiRs, 3' end modifications, and arm selection preferences of pre-miRNAs.