Several C. elegans cosmids and a subset of the Sprinzl tRNA database (animal cytoplasmic + eubacterial) are used in these examples to illustrate various features of the program; all have been included in the /Demo subdirectory so the user may also try out these examples. These files are written in the FASTA sequence format and have the file extension '.fa'. Run times are given for an R4400-200 Indigo SGI, and are approx. equal on a DEC Alpha 2100/400 190MHZ.
> tRNAscan-SE
> tRNAscan-SE F22B7.fa
The following (selected) run options will be printed first:
------------------------------------------------------------ Sequence file(s) to search: F22B7.fa Results written to: Standard output Output format: Tabular Searching with: tRNAscan + EufindtRNA -> Cove tRNAscan parameters: Strict EufindtRNA parameters: Relaxed (Int Cutoff= -32.1) ------------------------------------------------------------
This search produces tabular output results of 5 tRNAs for the 40kbp cosmid F22B7. Takes about 15 seconds.
> tRNAscan-SE -o mytrnas -m mystats C28G1.fa
The following (selected) run options will be printed first:
------------------------------------------------------------ Sequence file(s) to search: C28G1.fa Results written to: mytrnas Output format: Tabular Searching with: tRNAscan + EufindtRNA -> Cove tRNAscan parameters: Strict EufindtRNA parameters: Relaxed (Int Cutoff= -32.1) Search statistics saved in: mystats ------------------------------------------------------------
Default search mode, saves tabular output results for cosmid C28G1 in file 'mytrnas' and run statistics in file 'mystats'.
> tRNAscan-SE -a -o# -m# C28G1.fa
------------------------------------------------------------ Sequence file(s) to search: C28G1.fa Results written to: C28G1.out Output format: ACeDB Searching with: tRNAscan + EufindtRNA -> Cove tRNAscan parameters: Strict EufindtRNA parameters: Relaxed (Int Cutoff= -32.1) Search statistics saved in: C28G1.stats ------------------------------------------------------------
This is the same search as the immediately preceding example, but final results are saved in the file 'C28G1.out' and the statistics summary is saved in 'C28G1.stats' using the '#' shorthand for default output file names. Also, the results written to 'C28G1.out' are in the ACeDB format because the -a option was used.
> tRNAscan-SE -p mycosmid -f# -m# C28G1.fa
------------------------------------------------------------
Sequence file(s) to search: C28G1.fa
Results written to: Standard output
Output format: Tabular
Searching with: tRNAscan + EufindtRNA -> Cove
tRNAscan parameters: Strict
EufindtRNA parameters: Relaxed (Int Cutoff= -32.1)
tRNA secondary structure
predictions saved to: mycosmid.ss
Search statistics saved in: mycosmid.stats
------------------------------------------------------------
This search sends the tabular results to standard output since no -o option was used. Secondary structure information for all tRNAs was saved in file 'mycosmid.ss' and run stats were saved in 'mycosmid.stats' since the default file name prefix was specified as 'mycosmid' using the -p option.
> tRNAscan-SE -d -P -o# -n 'DE*' Sprz-sub.fa
------------------------------------------------------------ Sequence file(s) to search: Sprz-sub.fa Search only names matching: DE* Results written to: Sprz-sub.out Output format: Tabular Searching with: tRNAscan + EufindtRNA -> Cove tRNAscan parameters: Strict EufindtRNA parameters: Relaxed (Int Cutoff= -36) ------------------------------------------------------------
In this example, the program's progress is displayed (-d option) as it
searches through the input sequence file in this example. This search
will only analyze the 39 sequences in the file Sprz-sub.fa (1013 total
seqs) with names matching the 'DE*' key (ie. DE1140, DE1180, DE1200,
etc).
Since many of the sequences in this search are from prokaryotes, the
-P parameter is used to increase search sensitivity to detect
prokaryotic tRNAs that match the consensus A and B boxes less closely
than eukaryotic tRNAs (the EufindtRNA intermediate Cutoff is set to
-36 instead of the default -32.1; the lower the cutoff, the more
sensitive the search). Also, the prokaryotic tRNA model is used in
covariance model analysis.
> tRNAscan-SE -C -p F22B7cov -o# -m# F22B7.fa
------------------------------------------------------------ Sequence file(s) to search: F22B7.fa Results written to: F22B7cov.out Output format: Tabular Searching with: Cove only Search statistics saved in: F22B7cov.stats ------------------------------------------------------------
For users with the computational resources to spare, sequences can be analyzed using Cove analysis only (tRNAscan and EufindtRNA are not used as a pre-filter). This option is up to 3,000 times slower than using the default search mode, but may detect up to 1% of the tRNAs missed by tRNAscan-SE in default run mode. This example searches the F22B7 cosmid and finds no additional tRNAs over default tRNAscan-SE run mode. Takes about seventy minutes.
> tRNAscan-SE -T -p sprinzl.tscan -l# -m# Sprz-sub.fa
------------------------------------------------------------ Sequence file(s) to search: Sprz-sub.fa Results written to: Standard output Output format: Tabular Searching with: tRNAscan only tRNAscan parameters: Strict Search log saved in: sprinzl.tscan.log Search statistics saved in: sprinzl.tscan.stats ------------------------------------------------------------
Using the -T option approximates using tRNAscan v1.3 only (less sensitive and less selective). The -l option saves a log of the program run in 'sprinzl.tscan.log'. The search finds 962/1013 tRNAs in about 40 seconds, versus the same search in default mode which finds 1007/1013 tRNAs in about 17 minutes (search speed goes down for default mode as density of tRNAs in input sequence(s) goes up).
First, re-run an example from above, this time saving the tabular
output results:
> tRNAscan-SE -o# F22B7.fa
Now, suppose you'd like to see the secondary structures (and/or
produce ACeDB output) for the tRNAs in this sequence. Instead of
re-scanning the entire sequence with a completely new run, specify
that tRNAscan-SE should only scan those regions known to have tRNAs
from a previous tRNAscan-SE run. In the following example, the
pre-existing result file ``F22B7.out'' is specified with the -u
parameter, producing results faster than a complete re-scan of the
whole cosmid:
> tRNAscan-SE -u F22B7.out -f# -a -o F22B7.ace F22B7.fa
------------------------------------------------------------
Sequence file(s) to search: F22B7.fa
Results written to: F22B7.ace
Output format: ACeDB
Searching with: Cove only
Using previous
tabular output file: F22B7.out
tRNA secondary structure
predictions saved to: F22B7.ss
------------------------------------------------------------
The secondary structures for these tRNAs are saved in ``F22B7.ss'' and the results are saved in ACeDB format in the file ``F22B7.ace''. Useful for re-scanning just tRNAs in large sequences or sequence sets.
> tRNAscan-SE -g gcode.cilnuc DQ6060.fa
------------------------------------------------------------ Sequence file(s) to search: DQ6060.fa Results written to: Standard output Output format: Tabular Searching with: tRNAscan + EufindtRNA -> Cove tRNAscan parameters: Strict EufindtRNA parameters: Relaxed (Int Cutoff= -32.1) Alternate transl code used: from file gcode.cilnuc ------------------------------------------------------------
The tRNA used in this example is from Tetrahymena thermophila, a ciliate protozoan with an alternate translation of the codon TAA (normally a stop codon) to glutamine. By using the alternate translations specified in the included file ``gcode.cilnuc'', the correct tRNA type is output.