Reproduce sTCW overview

This describes how to obtain the table of results corresponding to statistics in the overview. The following short-hand is used:
  • The "Column:x" indicates that x should be selected for viewing in the table.
  • #Seqs is the number of sequences, which is listed at the top of the overview.
  • "Stats" is the "Show Column Stats" on the "Table..." drop-down.
Always clear filters before setting new ones!


Most of the input section is data supplied by the user with runSingleTCW. The following are computed:
SIZEShow AllColumn:Counts for all conditions: Stats, column:Sum
AVG-lenShow AllColumn:Length; Stats, column:Average
MED-lenShow AllColumn:Length; Stats, column:Median
The median in the two cases may be slightly different because they are computed differently.


Hit Statistics:

ColumnSearchObtain number
Sequences with hitsFilters: AnnotatedNumber of rows
Unique hitsAnnoDB Hits: Seq:None(slow)Hits # above table
Total sequence hitsAnnoDB Hits: Seq:None(slow)Pairs # above table
Bases covered by hitAnnoDB Hits: Seq:Best BitsUnselect "Group by Hit ID""; column:Align;
Stats, column:Sum; for NT, multiply by 3
Total bases
(residues for AA seqs)
Show AllColumn:Length; Stats, column:Sum


ColumnSearchObtain number
ONLY Filters: Annotated, Best Bits,
Enter DBtype and Taxonomy
    values for AnnoDB,
General <=1 annoDB
Number of rows
The following all use AnnoDB Hits panel with the correct ANNODB selected from the AnnoDBs panel.
BITSSeq:Best BitsSeqs # above table
ANNOSeq:Best AnnoSeqs # above table
UNIQUESeq:None(slow)Hits # above table
TOTALSeq:None(slow) Pairs # above table
AVG %SIMSeq:None(slow)Unselect "Group by Hit ID"; column:%Sim; Stats, column:Average
Rank=1 is the best hit for a sequence for a given annoDB.
HAS HITSeq:Rank=1Seqs # above table; percentage of total #Seqs
AVG %SIMSeq:Rank=1 Uncheck "Group by Hit"; Column:%Sim; Stats, column:Average
Cover >=NSeq:Rank=1,%Sim>=N,%HitCov*>=N Seqs # above table; percentage of HAS HIT
*HitCov is the difference between the hit stop and start coordinates divided by the length of the protein.

Top 15 species from total: N

The N is the number of unique species based on the first two words of the species name. From "AnnoDB Hits":
  • Select "Species"", select "Two words"", enter first two words of species name next to "Find", select "Find", select the entry on the left and add to the right.
  • Select "Best Bits", "Best Anno" or "None" for the three numbers shown.
  • Use the number listed beside "Pairs".

Gene Ontology Statistics:

ColumnSearchObtain number
Unique GOs GO Annotation: no filters Results number
Unique hits with GOs AnnoDB Hits: Seqs:None; GO,etc:Has GO Hits number at top of table
Sequences with GOs Filters: Best Hit: Annotated, Best with GO Number of rows
Seq best hit has GOs AnnoDB Hits: Seq:Best Bits; GO,etc:Has GO Seqs number at top of table
biological_process GO Annotation: Level: biological_process Number GOs at top of table
molecular_function GO Annotation: Level: molecular_function Number GOs at top of table
cellular_component GO Annotation: Level: cellular_component Number GOs at top of table
is_a, part_of GO Annotation: no filters Export..., Each GO's parents with relations, grep (see footnote*)
* From terminal, 'grep is_a GOeachParents.tsv | wc'. Repeat with is_a replaced with part_of.


The following sections may not exist if the input had no count files or the DE methods were not executed.
TPM* Filter: select Condition under Exclude; set "At Most" to 1.99. This will continue using 4.99, 9.99, etc, where the previous results are subtracted from the current. The results are for intervals >=N to <M.
Differential Expression Filter: select the condition pair, then enter the number, e.g. 1E-5. These counts are accumulative.
GO Enrichment: GO Annotation; select the condition pair using the Enrich panel; enter the threshold, e.g. 0.05. These counts are accumulative.
* If "RPKM" is at the top of the Overview instead of "TPM", then RPKM was computed instead of TPM.


If the sequences (e.g. ESTs) have been assembled, there are statistics on buried, mate-pairs, etc. Most of them are columns, so select the column, then use the "Show column stats" to view the sum. The following will only cover the three sections that are found in un-assembled sequences.
Sequence lengths Filters: General: change >= to <=, enter upper limit into Length The counts of the intervals to the left need to be subtracted from the number of rows.
ORF lengths Filter: SNPs and ORFs; enter lower limit into ORF len Work from right to left, the counts of the intervals to the right need to be subtracted from the number of rows.
Quality Filters: General: Has Ns: Yes; Column: Ns The resulting rows are the #n>0; sort the rows in descending order and scroll down to the row number that start having #n>10.

ORF Stats:

ColumnSearchObtain number
The following use the Basic Sequence. Select "TCW" and enter the Substring indicated.
Is Longest ORF!Lg#Seqs minus Result number
Markov Best Score!Mk#Seqs minus Result number
All of the above$Result number
ORF=HitORF=HitResult number
ORF=Hit with EndsORF=Hit+Result number
Multi-frameMulti-frameResult number
Stops in HitStopResult number
>=9 Ns in ORF9NResult number.
The following use the Filters: SNPs and ORFs.
Both EndsHas Ends (Start&Stop codons)Number of rows
Has HitProtein confirmationNumber of rows
ORF>=300Has ORF >= 300Number of rows

GC Content:

The only number reproducible is the GC Content, which is the %GC over the entire sequence.
GC ContentShow AllColumn:%GC; Stats, column:Average
Note, there will be some slight difference in the number due to round-off error.
  • The "Pos i" column is the percent of G or C in the ith positon of the CDS codons.
  • The %GC is the percent of G's and C's over the sequence length.
  • The CpG-O/E is ratio observed/expected [(#CpG/(#G*#C))*length].
  • The UTRs can be viewed in the Sequence Detail alignments, but there is no column for it.