# Tblastn Vdb

> Use when searching protein queries against translated SRA or WGS-backed VDB databases with BLAST.

- Skill: `vimalinx/tblastn-vdb` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add vimalinx/tblastn-vdb`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vimalinx/tblastn-vdb/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: vimalinx (https://skillmd.com/u/vimalinx)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/vimalinx/tblastn-vdb

---


# tblastn-vdb

## Quick Start

- **Command:** `tblastn_vdb -query proteins.fa -db <SRA_or_WGS_name> [options]`
- **Local executable:** `/home/vimalinx/miniforge3/envs/bio/bin/tblastn_vdb`
- **Full reference:** See [references/help.md](references/help.md)

## When To Use This Tool

- Search protein queries against translated nucleotide content exposed through SRA or WGS-backed VDB sources.
- Run `tblastn`-style searches without first building a local translated BLAST database.
- Control whether you search unaligned reads, aligned reference sequences, or both via `-sra_mode`.
- Reuse BLAST protein-query controls such as `-matrix`, `-evalue`, `-comp_based_stats`, or `-in_pssm`.

## Common Patterns

```bash
# 1) Search unaligned reads only with tabular output
tblastn_vdb \
  -query proteins.fa \
  -db SRR123456 \
  -sra_mode 0 \
  -outfmt 6 \
  -evalue 1e-6 \
  -num_threads 8
```

```bash
# 2) Search aligned reference sequences only
tblastn_vdb \
  -query proteins.fa \
  -db SRR123456 \
  -sra_mode 1 \
  -out aligned_refs.txt
```

```bash
# 3) Include filtered reads and search from a PSI-BLAST checkpoint
tblastn_vdb \
  -in_pssm profile.chk \
  -db SRR123456 \
  -sra_mode 2 \
  -include_filtered_reads \
  -outfmt "6 qaccver saccver pident length evalue bitscore"
```

## Recommended Workflow

1. Confirm the target is an SRA or WGS-backed VDB source and not an ordinary local BLAST database.
2. Decide whether you want unaligned reads, aligned reference sequences, or both, then set `-sra_mode` explicitly.
3. Choose reproducible search settings up front: `-outfmt`, `-evalue`, `-matrix`, `-comp_based_stats`, and `-num_threads`.
4. Start with a small query or accession subset before scaling the search across larger read-backed databases.

## Guardrails

- `-db` here means an SRA or WGS database name; it is not a standard local BLAST DB path.
- This build uses BLAST+ single-dash conventions such as `-help` and `-version`; autogenerated `--help` or `--version` captures are wrong for the live binary.
- Bare invocation fails with `Must specify at least one SRA/WGS database`, so missing `-db` is the first thing to check.
- `-sra_mode 0` searches unaligned reads only, `1` searches aligned reference sequences only, and `2` searches both; record that choice in reproducible workflows.
- `-query` and `-in_pssm` are mutually exclusive, and `-db_gencode` matters if the subject database needs a non-standard translation table.