Best way to go from long GO list to GO slim terms?

[sr320]

What is the best way to go from a file like

TRINITY_DN0_c0_g1	GO:0003674,GO:0003824,GO:0003964,GO:0006139,GO:0006259,GO:0006310,GO:0006313,GO:0006725,GO:0006807,GO:0008150,GO:0008152,GO:0009987,GO:0016740,GO:0016772,GO:0016779,GO:0032196,GO:0034061,GO:0034641,GO:0043170,GO:0044237,GO:0044238,GO:0044260,GO:0044699,GO:0044710,GO:0044763,GO:0046483,GO:0071704,GO:0090304,GO:1901360
TRINITY_DN0_c10_g1	GO:0003674,GO:0003824,GO:0004659,GO:0004660,GO:0005488,GO:0005575,GO:0005829,GO:0005875,GO:0005965,GO:0006464,GO:0006807,GO:0008150,GO:0008152,GO:0008270,GO:0008318,GO:0009987,GO:0016740,GO:0016765,GO:0018342,GO:0018343,GO:0019538,GO:0032991,GO:0036211,GO:0043167,GO:0043169,GO:0043170,GO:0043234,GO:0043412,GO:0044237,GO:0044238,GO:0044260,GO:0044267,GO:0044422,GO:0044424,GO:0044430,GO:0044444,GO:0044446,GO:0044464,GO:0046872,GO:0046914,GO:0071704,GO:0097354,GO:1901564,GO:1902494,GO:1990234
TRINITY_DN0_c2_g4	GO:0000166,GO:0003674,GO:0005488,GO:0005524,GO:0005575,GO:0005737,GO:0005856,GO:0017076,GO:0030554,GO:0032553,GO:0032555,GO:0032559,GO:0035639,GO:0036094,GO:0043167,GO:0043168,GO:0043226,GO:0043228,GO:0043229,GO:0043232,GO:0044424,GO:0044464,GO:0097159,GO:0097367,GO:1901265,GO:1901363

specifically - https://gannet.fish.washington.edu/Atumefaciens/20210318_cbai_trinotate_transcriptome-v4.0/20210318.cbai_transcriptome_v4.0.fasta.trinotate.go_annotations.txt

to a table of GO slim terms that could be easily grepped?

[sr320]

Good question! but have you every thought about the reverse. Say you wanted all genes related to Immune Response? why not just search for GO:0006955?
see http://amigo.geneontology.org/amigo/term/GO:0006955

[kubu4]

To address your first question:

awk '{print $2}' steven_GO.txt | tr "," "\n"

That will pull GO terms from input file (steven_GO.txt) and then create a newline-delimited file of all the GO terms.

Then, use GSEAbase in R to map to Biological Process GOslims (note: Requires goslim_generic.obo from http://geneontology.org/docs/go-subset-guide/) :

# Load necessary libraries
library(GSEABase)
library(tidyverse)

# Expects GO terms to be in first column of input file

## Get max number of fields
# Needed to handle reading in file with different number of columns in each row
max_fields <- max(na.omit((count.fields(item, sep = "\t", blank.lines.skip = TRUE))))
  
## Read in tab-delimited GOseq file
# Use "max_fields" to populate all columns with a sequentially numbered header
 go_seqs <- read.table(item,
                        sep = "\t",
                        header = TRUE,
                        col.names = paste0("V",seq_len(max_fields)),
                        fill = TRUE)
  
## Grab just the individual GO terms from the "category" column)
 goterms <- as.character(go_seqs$V1)
  
### Use GSEA to map GO terms to GOslims
  
## Store goterms as GSEA object
myCollection <- GOCollection(goterms)
  
## Use generic GOslim file to create a GOslim collection
  
# I downloaded goslim_generic.obo from http://geneontology.org/docs/go-subset-guide/
# then i moved it to the R library for GSEABase in the extdata folder
# in addition to using the command here - I think they're both required.
slim <- getOBOCollection("~/data/goslim_generic.obo")
  
## Map GO terms to GOslims and select Biological Processes group
slims <- goSlim(myCollection, slim, "BP", verbose = TRUE)

EDITED: Forgot to address GO to GOslim part of question.