# BLAST Tabular Output Guide

*By Belmin Bajramovic [@B-Bajramovic](https://github.com/B-Bajramovic)*

## 1. What is BLAST tabular output

BLAST supports **tabular output** via the `-outfmt` option.

The most commonly used formats are:

- `-outfmt 6` : tabular output, no comment lines
- `-outfmt 7` : tabular output, with comment lines starting with `#`

Tabular output is designed for:
- downstream parsing (awk, pandas, R)
- filtering and QC
- large-scale homology searches

## 2. The classic 12-column BLAST table

These are the historical default columns returned by `-outfmt 6`
when no custom field list is provided.

| #  | Field     | Description |
|----|-----------|-------------|
| 1  | qseqid    | Query sequence ID |
| 2  | sseqid    | Subject (target) sequence ID |
| 3  | pident    | Percentage of identical matches |
| 4  | length    | Alignment length |
| 5  | mismatch  | Number of mismatches |
| 6  | gapopen   | Number of gap openings |
| 7  | qstart    | Start of alignment in query |
| 8  | qend      | End of alignment in query |
| 9  | sstart    | Start of alignment in subject |
| 10 | send      | End of alignment in subject |
| 11 | evalue    | Expect value |
| 12 | bitscore  | Bit score |

This set is **not exhaustive** and is rarely optimal on its own.

## 3. Custom tabular output (recommended)

You can explicitly choose which columns BLAST writes:

```sh
-outfmt "6 qseqid sacc pident length qcovs evalue bitscore stitle"
```

## 4. Core blastp-specific fields

### Identifiers and annotation

- `qseqid`   : query sequence ID
- `sseqid`   : subject sequence ID
- `sacc`     : subject accession
- `saccver`  : subject accession.version
- `stitle`   : subject description line

### Alignment quality

- `pident`   : percent identity
- `length`   : alignment length
- `nident`   : number of identical residues
- `positive` : number of positive-scoring residues
- `gaps`     : total gaps
- `gapopen`  : gap openings

### Coordinates

- `qstart`, `qend`
- `sstart`, `send`

### Statistics

- `evalue`
- `bitscore`

## 5. Coverage and length fields (important)

Coverage fields are essential for filtering weak or spurious hits.

- `qlen`     : full query length
- `slen`     : full subject length
- `qcovs`    : query coverage per subject
- `qcovhsp`  : query coverage per HSP

Example:

```sh
-outfmt "6 qseqid sacc pident length qlen qcovs evalue bitscore"
```

## 6. Taxonomy-aware fields (nr database)

These fields are especially useful when searching the NCBI nr database.

- `ssciname`   : subject scientific name
- `scomname`   : subject common name
- `staxids`    : NCBI taxonomy ID(s)
- `sskingdom`  : superkingdom

Example:

```sh
-outfmt "6 qseqid sacc pident length qcovs evalue bitscore ssciname staxids"
```

## 7. Sequence-including fields (can inflate output)

These fields massively increase output size.

- `qseq`  : aligned query sequence
- `sseq`  : aligned subject sequence

Only use when manual alignment inspection is required.

## 8. Fields mainly for nucleotide-based BLAST

Relevant for blastn, tblastn, blastx, tblastx:

- `qframe`  : query frame
- `sframe`  : subject frame
- `sstrand` : subject strand

Irrelevant for blastp.

## 9. Recommended column sets

### General blastp annotation

```sh
-outfmt "6 qseqid sacc pident length qcovs evalue bitscore stitle"
```

### Strict homology filtering

```sh
-outfmt "6 qseqid sacc pident length nident qlen qcovs evalue bitscore"
```

### Taxonomy-aware screening

```sh
-outfmt "6 qseqid sacc pident length qcovs evalue bitscore ssciname staxids"
```

### Alignment inspection / debugging

```sh
-outfmt "6 qseqid sacc qstart qend sstart send pident length qseq sseq"
```

## 10. Other BLAST programs (summary)

| Program  | Query type | Subject type | Notes |
|---------|------------|--------------|-------|
| blastp  | protein    | protein      | Most common for proteomics |
| blastn  | nucleotide | nucleotide   | Strand and frame fields relevant |
| blastx  | nucleotide | protein      | Query translated |
| tblastn | protein    | nucleotide   | Subject translated |
| tblastx | nucleotide | nucleotide   | Both translated |

Most fields are shared, but frame/strand fields only apply to nucleotide
or translated searches.

## 11. How to list all available fields locally

Run:

```sh
blastp -help
```

Search for the section titled **Custom output formats**.
The exact field list may vary slightly between BLAST+ versions.

## 12. Final advice

- Always specify your `-outfmt` fields explicitly
- Always include at least one coverage metric
- Avoid sequence fields unless necessary
- Keep column order stable for downstream scripts