RNA-Seq is an expensive and time-consuming procedure, as it requires the preparation of an entire genomic library. Another limitation is the difficulty in accurately estimating gene expression. Small transcripts may be more difficult to count due to the standard size selection of RNA-Seq libraries. Also, two different genes may sometimes have overlapping transcripts. Additionally, RNA-Seq can result in transcript-length bias due to the multiple fragmentation and cDNA or RNA size-selection steps they require. This bias may cause issues for downstream analysis. Another disadvantage is that the RNA-seq signal across transcripts tends to display non-uniformity of coverage, which may be caused from priming with random hexamers, cDNA synthesis, and ligation. One must also consider the read mapping uncertainty during the quantification of transcripts. Read mapping may be affected by sequencing error rates, incomplete genome sequence, and inaccuracies in transcript annotations. The normalization of the number of reads mapping to each transcript (based on transcript length) is also potentially a source of error in quantitative applications due to events such as alternative splicing. Another limitation is that when analyzing sRNAs, RNA-Seq may not display an absolutely quantitative view of these transcripts. The number of reads obtained per sRNA does not necessarily correlate with their true abundance. RNA-seq techniques also often involve a poly (A) mRNA enrichment step.