Decoding Ancient Decay: How Software Replicates the Chemical Degeneracy of Ancient Indian Genomes

When a biological sample is excavated from the warm, acidic soils of the Indian subcontinent, the genetic material within has already undergone centuries of chemical collapse. Over thousands of years, the long, tidy double helix of living DNA breaks down into short fragments. Water molecules slowly strip amino groups from cytosine bases, transforming them into uracil—a process called deamination.

When modern high-throughput sequencing machines read these damaged fragments, they misinterpret these chemical scars. The machine reads the post-mortem uracil as thymine, introducing systematic errors known as transition substitutions.

For geneticists analyzing ancient South Asian human remains, these chemical modifications present a serious problem. If a software algorithm cannot distinguish between an actual ancient genetic mutation and a spot of post-mortem chemical damage, it can produce false matches. This can lead to incorrect conclusions about ancient migration, population structure, and biological kinship.

To solve this problem, researchers require a clean, synthetic baseline. A newly released command-line software package named SCAR (Single-Cell and Ancient Reads simulator) allows geneticists to introduce controlled, highly customizable degradation patterns directly into pristine modern sequences. Developed by researchers and published on bioRxiv (SCAR: Controlled mutations, ancient DNA damage, and fragmentation of fasta and fastq sequences), the tool gives scientists a way to stress-test their analytical pipelines by simulating the precise chemical geometry of decay.

The Chemistry of Post-Mortem Deamination

The physical destruction of ancient DNA follows a predictable spatial pattern. Within a buried bone, environmental moisture drives hydrolytic damage. This damage is most severe at the exposed single-stranded ends of the degrading DNA fragments.

As cytosine loses its amine group and becomes uracil, the sequencing machinery reads it as a thymine (T) instead of a cytosine (C). On the complementary strand, guanine (G) is read as adenine (A). This bias causes a characteristic spike in C-to-T transitions at the 5-prime (the beginning of the DNA strand) terminus, and G-to-A transitions at the 3-prime (the end of the DNA strand) terminus.

These terminal substitutions are the molecular signature of genuine ancient DNA. While these chemical markers help confirm that a sample is truly ancient—rather than modern laboratory contamination—they simultaneously degrade the quality of the sequence.

In South Asian contexts, where high ambient heat and seasonal monsoons accelerate chemical decay, recovered fragments are often extremely short, sometimes averaging fewer than 40 base pairs in length. When these short, highly damaged sequences are aligned to a modern human reference genome, the combination of short fragment length and chemical substitutions can cause the sequences to map to the wrong genomic regions.

How SCAR Replicates the Soil

SCAR is written in C++ (SCAR bioRxiv). It allows researchers to take clean, high-coverage genetic sequences from modern populations and systematically break them down to match the exact preservation state of an ancient sample.

The software operates through three primary, independent modules: fragmentation, mutation, and ancient DNA damage (SCAR bioRxiv).

1. Fragmentation

Instead of simply shearing DNA at random, SCAR allows users to define custom fragment-length distributions. Researchers can model this using mathematical curves, such as lognormal or exponential distributions, or by importing empirical fragment-length tables directly from real archaeological datasets (SCAR bioRxiv). This replication is critical because fragment length directly dictates how uniquely a sequence can be mapped to the human genome.

2. Mutation

To simulate evolutionary divergence over time, SCAR can introduce specific transition and transversion mutations at defined rates across the entire sequence (SCAR bioRxiv). This allows researchers to model how ancient lineages diverged from ancestral populations.

3. Position-Specific Damage

This module recreates the chemical deamination curve. Rather than applying a flat error rate across the sequence, SCAR uses position-specific probability tables (SCAR bioRxiv). The probability of a C-to-T substitution is highest at the absolute terminal base of the fragment and decreases exponentially as it moves toward the center of the molecule. Users can import these damage curves directly from mapDamage, an industry-standard software tool used to track post-mortem damage in real ancient samples (SCAR bioRxiv).

By running these steps in sequence—first fragmenting the DNA, then introducing mutations, and finally applying terminal damage—SCAR ensures that the synthetic chemical decay is physically concentrated on the newly created ends of the short fragments, exactly mimicking natural preservation (SCAR bioRxiv).

Preventing False Matches in South Asian Genomics

The practical value of this simulation is significant for South Asian paleogenomics. Ancient DNA recovered from the region is exceptionally rare due to poor preservation environments. When a sample is successfully sequenced, the data yield is often low-coverage, meaning many positions in the genome are read only once or twice.

In low-coverage genomes, researchers cannot easily verify whether a mutation is real or just chemical damage by comparing multiple overlapping reads. If a damaged site undergoes a C-to-T transition, the software might interpret this as a heterozygous site—a position where an individual inherited different genetic variants from each parent.

This misidentification can skew calculations of heterozygosity, which geneticists use to estimate ancient population sizes, levels of inbreeding, and past demographic bottlenecks. During validation trials, SCAR’s developers demonstrated that uncorrected post-mortem damage artificially inflates calculated heterozygosity rates, creating false signals of genetic diversity (SCAR bioRxiv).

By generating synthetic datasets with known, controlled levels of deamination and fragmentation, geneticists can run these simulated sequences through their variant-calling pipelines. Since the true sequence is known beforehand, researchers can measure exactly how many false variants are introduced by the damage. This allows them to calibrate their filtering software to ignore deaminated bases at the ends of reads, preserving the accuracy of the underlying genetic data without throwing away valuable sequence information.

As paleogenomics continues to study the demographic history of South Asia, tools like SCAR provide a vital digital sandbox. By mastering the mathematics of decay, researchers can ensure that the ancient stories written in damaged DNA are read exactly as they were left in the soil.

Peer-Reviewed Sources:

• Hartmann, M. et al. (2026). "SCAR: Controlled mutations, ancient DNA damage, and fragmentation of fasta and fastq sequences." bioRxiv. https://www.biorxiv.org/content/10.64898/2026.09.24.754018v1