Target Prediction Algorithms and Bioinformatics Resources for miRNA Studies - MicroRNAs as Tools in Biopharmaceutical Production - page 43

Table 3.2 Summary of online data repositories for microRNAs. The nature of the data included,
source of miRNA target predictions (if present) and web address of resources are presented below
Database Data type Prediction

algorithm

Analytical
techniques

URL

miRBase Sequence NA NA http://www.mirbase.org/

microRNA.org Computational
prediction

miRanda
miRSVR

NA www.microRNA.org

microCosm Computational

prediction

miRanda NA http://www.ebi.ac.uk/

enright-srv/
microcosm/htdocs/
targets/v5/

Tarbase Confirmed targets

Expression profiling

NA Luciferase,

Microar-
ray, NGS,
Proteomics

http://www.microrna.gr/
tarbase

MirTarbase Construct validation

Expression profiling
data

NA Microarray,

NGS,
Proteomics

http://mirtarbase.mbc.
nctu.edu.tw/

starBase microRNA-target

interactions,
protein-RNA
interaction,
Degradome
sequences

NA HTS-CLIP,

Degradome-
Seq

http://starbase.sysu.
edu.cn/

miRecords Computational and

Confirmed Targets

TargetScan,
PicTar, PITA,
RNAhybrid,
DIANA-microT,
RNA22

Luciferase,
Microar-
ray, NGS,
Proteomics

http://mirecords.biolead.
org/doc.php

miRNA and targets gene expression at specific phases of culture was determined to
further prioritise targets and enrichment of biological processes relating to the stages
growth rate (e.g. cell proliferation, cell cycle and apoptosis) was observed for groups
of targets inversely related to their corresponding miRNAs.
Proteomic profiling has also been carried out in CHO cells with artificially overex-
pressed miR7 (Meleady et al. 2012 ), known to inhibit CHO cell growth and increase
productivity (Barron et al. 2011 ). Quantitative label-free mass spectrometry was
utilised to measure the differences between negative control cells and cells transfected
with miR7 pre-mirs, with a total of 93 and 74 proteins found to be downregu-
lated and upregulated respectively. Potential targets of those downregulated proteins
were identified through the miRWalk interface by combining the outputs of six tar-
get prediction algorithms (miRanda, miRwalk, miRDB, RNA22, RNAhybrid and
TargetScan). The miRWalk system was configured to identify target matches with a
minimum seed length of seven in the 5 UTR, 3 UTR, promoter or CDS. Only targets
with a p-value of < 0.05 were returned and targets were deemed significant if they
were identified by two or more algorithms in either human, mouse or rat. The analysis
identified catalase (CAT) and stathmin (STMN1) as potential direct targets of miR7