explotiv de novo transcriptome explorer

Mar 16, 2016

Camille Scott

Lab for Data Intensive Biology

UC Davis

Sample Sunburst Plots

explotiv's primary output
These convey the "affinity" of a collection of transcripts for all clades in a base tree
Color is scaled 0.0-1.0 based on the affinity score

Calculating Scores

We start with a mapping between transcripts and proteins
Each transcript will likely map to many proteins
The protein databaes are currated, and provide a species mapping



In [1]:

    
from dammit.fileio import maf
import pandas as pd

aln_df = pd.concat(maf.MafParser('/Users/camille/w/scratch/explotiv-data/lamprey.500.fa.dammit/lamprey.500.fa.x.orthodb.maf'))



In [3]:

    
aln_df.sort_values('q_name').head()









    Out[3]:






  
    
      
      E
      EG2
      q_aln_len
      q_len
      q_name
      q_start
      q_strand
      s_aln_len
      s_len
      s_name
      s_start
      s_strand
      score
      bitscore
    
  
  
    
      700
      2.000000e-32
      3.600000e-26
      288
      528
      Transcript_102
      4
      +
      96
      296
      ASIS016320-PA
      199
      +
      269
      144.316791
    
    
      717
      4.300000e-06
      2.600000e+00
      66
      528
      Transcript_102
      461
      -
      22
      388
      ADAC009551-PA
      0
      +
      106
      58.425095
    
    
      716
      2.800000e-31
      4.600000e-25
      297
      528
      Transcript_102
      230
      -
      106
      503
      AGAP003997-PB
      88
      +
      262
      140.628191
    
    
      715
      2.800000e-31
      4.600000e-25
      297
      528
      Transcript_102
      230
      -
      106
      435
      AFAF019329-PA
      14
      +
      262
      140.628191
    
    
      714
      4.800000e-36
      1.200000e-29
      294
      528
      Transcript_102
      230
      -
      99
      393
      TCOGS2:TC005606-PA
      11
      +
      291
      155.909535

Take the normalized alignment scores for each transcript and map them to their corresponding species
These will be leaf nodes
Then propogate these scores up the tree
Score for an internal node is (sum of scores of its descendants) / (sum of branch lengths)
- Inspired by existing methods for calculating phylogenetic signal
- Caveat: my tree is actually a taxonomy, and all branch lengths are 1.
Do this for every transcript to get a score distribution for each node

Rhinella marina, the cane toad

![example](Rhinella_marina.svg)

Seriola lalandi, a ray-finned fish

![example](Seriola_lalandi.svg)

The chickadee, a small bird / greatly diminshed dinosaur

![example](chickadee.svg)

Crella elegans, a sponge

![example](Crella_elegans.svg)

Homaru americanus, a lobster

![example](Homaru_americanus.svg)

species from Pomacea, group of freshwater snails

![example](Pomacea.svg)

A skink species in the genus Carlia

![example](Carlia_N.svg)

Bactrocera tryona, the queensland fruit fly

![example](Bactrocera_tryona.svg)

Montastraea cavernosa, a coral

![example](Montastraea_cavernosa.svg)

Campylomormyrus compressirostris, an electric fish

![example](Campylomormyrus_compressirostris.svg)

Squalius pyrenaicus, a freshwater fish

![example](Squalius_pyrenaicus.svg)

Petromyzon marinus, the sea lamprey

![example](lamp10.svg)

The original plan included a complete annotation browser to be integrated with dammit
This was overambitious.
- Did a fair amount of work on it, but paused in favor of phylogenetic view
Integrating with existing software is more work than starting from scratch...

	E	EG2	q_aln_len	q_len	q_name	q_start	q_strand	s_aln_len	s_len	s_name	s_start	s_strand	score	bitscore
700	2.000000e-32	3.600000e-26	288	528	Transcript_102	4	+	96	296	ASIS016320-PA	199	+	269	144.316791
717	4.300000e-06	2.600000e+00	66	528	Transcript_102	461	-	22	388	ADAC009551-PA	0	+	106	58.425095
716	2.800000e-31	4.600000e-25	297	528	Transcript_102	230	-	106	503	AGAP003997-PB	88	+	262	140.628191
715	2.800000e-31	4.600000e-25	297	528	Transcript_102	230	-	106	435	AFAF019329-PA	14	+	262	140.628191
714	4.800000e-36	1.200000e-29	294	528	Transcript_102	230	-	99	393	TCOGS2:TC005606-PA	11	+	291	155.909535

explotiv de novo transcriptome explorer

Mar 16, 2016

Camille Scott

Lab for Data Intensive Biology

UC Davis

Sample Sunburst Plots

Calculating Scores

*Rhinella marina*, the cane toad

*Seriola lalandi*, a ray-finned fish

The chickadee, a small bird / greatly diminshed dinosaur

*Crella elegans*, a sponge

*Homaru americanus*, a lobster

species from *Pomacea*, group of freshwater snails

A skink species in the genus *Carlia*

*Bactrocera tryona*, the queensland fruit fly

*Montastraea cavernosa*, a coral

*Campylomormyrus compressirostris*, an electric fish

*Squalius pyrenaicus*, a freshwater fish

*Petromyzon marinus*, the sea lamprey