---
title: "blast"
output: html_document
date: "2023-03-30"
---

#Download Software
library(tidyverse)
library("kableExtra")
library("dplyr")

```{bash}
cd /home/jovyan/applications
curl -O https://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST/ncbi-blast-2.13.0+-x64-linux.tar.gz
```
```{bash}
#unzip the file
cd /home/jovyan/applications
tar -xf ncbi-blast-2.13.0+-x64-linux.tar.gz
```

```{bash}
~/applications/ncbi-blast-2.13.0+/bin/blastx -h
```
```{bash}
cd data
cd ../data
curl -O https://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz
mv uniprot_sprot.fasta.gz uniprot_sprot_r2023_01.fasta.gz
gunzip -k uniprot_sprot_r2023_01.fasta.gz
ls ../data
```
```{bash}
~/applications/ncbi-blast-2.13.0+/bin/makeblastdb \
-in ../data/uniprot_sprot_r2023_01.fasta \
-dbtype prot \
-out ../blastdb/uniprot_sprot_r2023_01
```
```{bash}
curl https://eagle.fish.washington.edu/cnidarian/Ab_4denovo_CLC6_a.fa \
-k \
> ../data/Ab_4denovo_CLC6_a.fa
```
```{bash}
head ../data/Ab_4denovo_CLC6_a.fa
echo "How many sequences are there?"
grep -c ">" ../data/Ab_4denovo_CLC6_a.fa
```
```{bash}
~/applications/ncbi-blast-2.13.0+/bin/blastx \
-query ../data/Ab_4denovo_CLC6_a.fa \
-db ../blastdb/uniprot_sprot_r2023_01 \
-out ../output/Ab_4-uniprot_blastx.tab \
-evalue 1E-20 \
-num_threads 20 \
-max_target_seqs 1 \
-outfmt 6
```
```{bash}
head -2 ../output/Ab_4-uniprot_blastx.tab
wc -l ../output/Ab_4-uniprot_blastx.tab
```

```{bash}
curl -o "uniprot_table_r2023_01.tab" -H "Accept: text/plain; format=tsv" "https://gannet.fish.washington.edu/seashell/snaps/uniprot_table_r2023_01.tab"
```
```{bash}
head -2 ../output/Ab_4-uniprot_blastx.tab
wc -l ../output/Ab_4-uniprot_blastx.tab
```

```{bash}
tr '|' '\t' < ../output/Ab_4-uniprot_blastx.tab | head -2
```
```{bash}
tr '|' '\t' < ../output/Ab_4-uniprot_blastx.tab \
> ../output/Ab_4-uniprot_blastx_sep.tab
```

#this chunk works but is not consistent with the assignement instructions (it uses a .fasta file instead 
# of a .tab)
```{bash}
head -2 ../data/uniprot_table_r2023_01.tab
wc -l ../data/uniprot_table_r2023_01.tab
```

```{r}
bltabl <- read.csv("../output/Ab_4-uniprot_blastx_sep.tab", sep = '\t', header = FALSE)
```

```{r}
spgo <- read.csv("../data/uniprot_table_r2023_01.tab", sep = '\t', header = TRUE)
```

```{r}
str(spgo)
```
```{r}
left_join(bltabl, spgo,  by = c("V3" = "Entry")) %>%
  select(V1, V3, V13, Protein.names, Organism, Gene.Ontology..biological.process., Gene.Ontology.IDs) %>% mutate(V1 = str_replace_all(V1, 
            pattern = "solid0078_20110412_FRAG_BC_WHITE_WHITE_F3_QV_SE_trimmed", replacement = "Ab")) %>%
  write_delim("../output/blast_annot_go.tab", delim = '\t')
```

```{r}
annot_tab <- read.csv("../output/blast_annot_go.tab", sep = '\t', header = TRUE)
```

```{r}
str(annot_tab)
```