Genome Profiling - Usage Guide

Overview

Genome profiling is the reference-free, pre-assembly step that turns raw sequencing reads into a k-mer frequency spectrum and models it to estimate the genome's size, heterozygosity, repeat content, and ploidy. Those four numbers are not trivia: they set the expected assembly size (the NG50 denominator and the haplotig sanity check), the genome-size parameter for assemblers like Flye, the homozygous-coverage and purge settings for hifiasm, and the up-front decision about whether short reads can even produce the assembly being requested. This skill counts k-mers with KMC or Jellyfish, models the spectrum with GenomeScope2, and infers ploidy independently with Smudgeplot, while flagging the common traps that make the estimate wrong: too-low coverage, noisy ONT reads, contamination spikes, a too-small k, and confusing k-mer coverage with sequencing coverage.

Prerequisites

Install the counters and modeling tools (Bioconda):

conda install -c bioconda kmc jellyfish genomescope2 smudgeplot merqury kat

• KMC or Jellyfish produces the k-mer histogram; only one is required.
• GenomeScope2 needs R; the Bioconda package pulls it in.
• Merqury provides best_k.sh (k selection) and meryl; $MERQURY should point at the install.
• Accurate reads are required for counting: Illumina or PacBio HiFi. Noisy ONT reads cannot be profiled reliably.
• A rough expected genome size (even an order of magnitude) helps choose k and validate the result.

Quick Start

Tell your AI agent what you want to do: - "Profile my genome from Illumina reads before I assemble" - "Estimate genome size and heterozygosity from these reads" - "Build a k-mer spectrum and run GenomeScope2" - "Figure out the ploidy of this non-model organism" - "Check whether my assembly is 2x too big because of haplotigs"

Example Prompts

Pre-assembly profiling

"I have 60x Illumina reads from an unknown beetle. Count 21-mers with KMC, run GenomeScope2 as diploid, and tell me the estimated genome size, heterozygosity, and repeat content, plus whether short reads will struggle with the heterozygosity."

Choosing k and counting

"My expected genome size is about 12 Mb. Use Merqury best_k.sh to pick k, count k-mers with that k, and run GenomeScope2 so I can set the genome-size parameter for Flye."

Ploidy inference

"GenomeScope2 fits poorly as diploid for this plant. Run Smudgeplot on the het k-mer pairs and tell me whether it looks triploid or tetraploid, and whether that disagrees with GenomeScope."

Diagnosing a bad spectrum

"The GenomeScope size estimate is twice what I expected and the spectrum has an extra peak. Help me decide whether this is contamination, an organelle spike, or a too-small k."

Sanity-checking an assembly

"My HiFi assembly is 1.8x the GenomeScope estimate and BUSCO Duplicated is high. Confirm whether this is uncollapsed haplotigs and what to do before reporting the genome size."

What the Agent Will Do

1. Confirm the reads are accurate (Illumina/HiFi) and adequate depth (homozygous peak >= ~15-20x); route noisy-ONT-only cases to assemble-then-estimate.
2. Choose k with Merqury best_k.sh from the expected genome size (k=21 for ~Gb genomes), using the same k for counting and modeling.
3. Count k-mers with KMC (or Jellyfish with -C), keeping singletons (-ci1) and capping the tail (-cs10000) to build the histogram.
4. Run GenomeScope2 with the matching -k and a starting ploidy -p, then read the model summary: haploid genome length, heterozygosity, and percent unique (repeat content).
5. Interpret the spectrum: single vs het (AB) peak, repeat tail, organelle/contamination spikes; distinguish k-mer coverage (lambda) from sequencing depth.
6. Run Smudgeplot to infer ploidy from het k-mer-pair coverage ratios and reconcile it with GenomeScope's -p.
7. Hand the estimate downstream: the NG50 denominator and haplotig sanity check (assembly-qc), Flye -g, hifiasm --hom-cov/purge level, and the short-vs-long-read decision.

Tips

• Always use the same k for counting and for GenomeScope2's -k; a mismatch silently mis-scales the model.
• Count canonical k-mers: KMC does this by default, Jellyfish needs -C. WGS is unstranded, so forward and reverse k-mers must be merged.
• Read the homozygous peak position off the plot rather than assuming it equals sequencing depth; they differ by the (L-k+1)/L factor.
• A distinct peak at roughly half the homozygous coverage is the heterozygous (AB) signal; its area scales with heterozygosity.
• A spike at very high multiplicity is usually organelle or high-copy repeat DNA, not the nuclear genome; cap the histogram tail so it does not dominate the fit.
• If GenomeScope's diploid fit is poor, treat ploidy as unknown and let Smudgeplot decide before forcing a -p.
• Never report an assembly size larger than the profiled estimate as the genome size; a 1.5-2x excess is almost always uncollapsed haplotigs.

Related Skills

• short-read-assembly - High het from the profile predicts short-read fragmentation and haplotig inflation
• hifi-assembly - The profiled size/het sets hifiasm --hom-cov and the purge level
• long-read-assembly - The profiled genome size feeds Flye -g and read-chemistry expectations
• assembly-qc - The estimate is the NG50 denominator and the 1.5-2x haplotig sanity check
• read-qc/quality-reports - QC and contamination-screen reads before profiling and assembling
• workflows/genome-assembly-pipeline - Profiling is the first step of the end-to-end assembly workflow