Soloutions Day 2:


2.1.1

awk '$3=="transcript"' SRR7889582.assemble.gtf | wc -l
>5019

awk '$3=="transcript"' SRR7889581.assemble.gtf | wc -l
>15378

awk '$3=="transcript"' merged.assemble.gtf | wc -l
>23550


2.1.2

awk '$4=="u"' gffcompare.tracking | wc -l
>2083


2.1.3

Do
grep '"u"' gffcompare.annotated.gtf | head
And use the coordinates of some unannotated transcripts to explore them at IGV


2.2.1

awk '$3=="transcript"' gffcompare.annotated.filtered.gtf |   wc -l
>3497
grep '"u"' gffcompare.annotated.filtered.gtf | wc -l
>39


3.1.1

From Whippet's github:
"In general, significant nodes should have a higher probability (e.g. >=0.90) suggesting there is likely to be a difference between the samples, and (2) significant nodes should have an absolute magnitude of change (|DeltaPsi| > x) where x is some value you consider to be a biologically relevant magnitude of change (we suggest |DeltaPsi| > 0.1)."

3.1.2

zcat Delta/neuron_vs_glia_new.diff.gz  | awk '($8>0.1||$8<0.1) && $9>=0.9' | cut -f 5 | sort | uniq -c
     12 AA
     21 AD
      8 AF
      4 AL
    107 CE
     42 RI
     86 TE
     69 TS
     1 Type
Thus, CE (cassette exons) are the most common type of alternative node between these conditions.