(c) Load the alignment files using the menu File, Load from
file or Load from URL depending on the case.
(d) Upload annotation files to show novel transcript isoforms
or alternative spliced transcripts.
10. To identify long noncoding RNAs, use the GTF file containing
the merged assemblies produced by Cuffmerge. Filter-out wellannotated genes in a spreadsheet editor and keep those that
have not been annotated as protein-coding genes in your reference genome. Include also those that have been annotated as
hypothetical proteins. Keep transcripts longer than 200 nt.
11. Predict lncRNA open reading frames (ORFs) using NCBI
Open Reading Frame Finder (www.ncbi.nlm.nih.gov/
orffinder/). NCBI ORF Finder searches for open reading
frames in the DNA sequence supplied by the user and returns
the range of each ORF, along with its protein translation. Enter
query sequence in FASTA format and choose searching parameters. Click Submit bottom. For sequences longer than
50 kb, a stand-alone version is available for Linux x64. Keep
only those transcripts that do not contain ORFs longer than
100 amino acids.
12. Perform the coding potential analysis of transcripts using two
different alignment-free methods:
(a) The support vector machine-based classifier, Coding
Potential Calculator 2 (CPC2) (http://cpc2.cbi.pku.
edu.cn). Enter query sequence in FASTA, BED, or GTF
format. When using a FASTA file as input, the reference
genome is not required. Select Genome assembly version
if your input file is in BED or GTF format. Click “Check
the reverse strand too”, and press bottom Run. The
CPC2 server uses RNA transcript sequences as input,
and outputs protein-coding probability of the transcripts
using four intrinsic features: Fickett score derived from
the weighted nucleotide frequency of the inputted fulllength transcript, and the rest of the three features, ORF
length, ORF integrity, and isoelectric point, which are
calculated based on the longest putative ORF identified
in silico. For lncRNAs, coding probability must be
between 0 and 1 and the resulting label must indicate
“noncoding.”
(b) The Coding Potential Assessment Tool (CPAT) (http://
lilab.research.bcm.edu/cpat/): Enter a query sequence in
FASTA or BED format. When using a FASTA file as input,
the reference genome is not required. Select Species
assembly if your input file is in BED format. Click Submit
button. CPAT uses RNA transcript sequences as input,
and outputs its coding probability by using logistic
466
Soledad Traubenik et al.
file or Load from URL depending on the case.
(d) Upload annotation files to show novel transcript isoforms
or alternative spliced transcripts.
10. To identify long noncoding RNAs, use the GTF file containing
the merged assemblies produced by Cuffmerge. Filter-out wellannotated genes in a spreadsheet editor and keep those that
have not been annotated as protein-coding genes in your reference genome. Include also those that have been annotated as
hypothetical proteins. Keep transcripts longer than 200 nt.
11. Predict lncRNA open reading frames (ORFs) using NCBI
Open Reading Frame Finder (www.ncbi.nlm.nih.gov/
orffinder/). NCBI ORF Finder searches for open reading
frames in the DNA sequence supplied by the user and returns
the range of each ORF, along with its protein translation. Enter
query sequence in FASTA format and choose searching parameters. Click Submit bottom. For sequences longer than
50 kb, a stand-alone version is available for Linux x64. Keep
only those transcripts that do not contain ORFs longer than
100 amino acids.
12. Perform the coding potential analysis of transcripts using two
different alignment-free methods:
(a) The support vector machine-based classifier, Coding
Potential Calculator 2 (CPC2) (http://cpc2.cbi.pku.
edu.cn). Enter query sequence in FASTA, BED, or GTF
format. When using a FASTA file as input, the reference
genome is not required. Select Genome assembly version
if your input file is in BED or GTF format. Click “Check
the reverse strand too”, and press bottom Run. The
CPC2 server uses RNA transcript sequences as input,
and outputs protein-coding probability of the transcripts
using four intrinsic features: Fickett score derived from
the weighted nucleotide frequency of the inputted fulllength transcript, and the rest of the three features, ORF
length, ORF integrity, and isoelectric point, which are
calculated based on the longest putative ORF identified
in silico. For lncRNAs, coding probability must be
between 0 and 1 and the resulting label must indicate
“noncoding.”
(b) The Coding Potential Assessment Tool (CPAT) (http://
lilab.research.bcm.edu/cpat/): Enter a query sequence in
FASTA or BED format. When using a FASTA file as input,
the reference genome is not required. Select Species
assembly if your input file is in BED format. Click Submit
button. CPAT uses RNA transcript sequences as input,
and outputs its coding probability by using logistic
466
Soledad Traubenik et al.