Disputed Paternity and Maternity: Forensic Testing in India

The starting point is statutory. Section 112 of the Indian Evidence Act 1872, now re-enacted nearly verbatim as Section 116 of the Bharatiya Sakshya Adhiniyam 2023, creates a strong rebuttable presumption: a child born during the continuance of a valid marriage between the mother and any man, or within 280 days after its dissolution (the mother remaining unmarried), is conclusive proof of legitimacy unless it is shown that the parties to the marriage had no access to each other at any time when the child could have been begotten.

That "no access" carve-out is the only door through which a court will admit DNA evidence to displace the presumption. Indian appellate courts have built a steady line of authority around it.

• Goutam Kundu vs State of West Bengal (1993). The Supreme Court held that courts in India cannot order blood (or DNA) tests as a matter of course. There must be a strong prima facie case, the test must be eminent need, and the consequences (bastardising a child) must be weighed.
• Bhabani Prasad Jena vs Convenor Secretary, Orissa State Commission for Women (2010). Reaffirmed Goutam Kundu and laid down the "eminent need" test: a DNA test is not a routine forensic tool but an exception, ordered only when no other evidence will resolve the dispute.
• Nandlal Wasudeo Badwaik vs Lata Nandlal Badwaik (2014). The Court accepted a DNA exclusion as overriding the Section 112 presumption where the husband had proved non-access. The case is the favourite NET citation because it confirms that a scientifically valid DNA exclusion will, in the right facts, defeat the statutory presumption.
• Aparna Ajinkya Firodia vs Ajinkya Arun Firodia (2023). Cautioned again against routine DNA orders in matrimonial disputes; the child's right to identity and dignity must be balanced against the petitioner's evidentiary need.

The standard examiner phrasing to remember: presumption of legitimacy is strong but rebuttable, the only statutory ground of rebuttal is non-access, and the Supreme Court has built a judicial filter (Goutam Kundu / Bhabani Prasad Jena / Aparna Firodia) on top of that statutory door to keep DNA orders rare and reasoned.

For criminal-side disputes (rape resulting in pregnancy, infanticide, child trafficking), the test is ordered under the inherent powers of the trial court read with the BSA 2023 expert-evidence regime; the chain of custody requirements are identical to any other biological exhibit.

Before DNA, Indian paternity disputes were settled (or, more often, not settled) by red-cell and serum protein typing. NTA still tests this chapter because it sets up the exclusion-vs-inclusion logic.

The classical workflow used three layers:

1. Red-cell antigens. ABO, Rh, MNS, Kell, Duffy, Kidd. ABO alone can resolve a clean exclusion in roughly 17 to 20 percent of falsely-accused men in an Indian population, because some father-child combinations are genetically impossible (an AB father cannot have an O child, an O father cannot have an AB child).
2. Red-cell enzymes. PGM, EsD, AcP, ADA, GLO. Each adds a few percent to the cumulative exclusion power.
3. Serum proteins by isoelectric focusing. Haptoglobin (Hp), group-specific component (Gc), transferrin (Tf), C3. These were the high-resolution end of the pre-DNA era.

Even a full panel of red-cell antigens, enzymes and serum proteins could only push the cumulative power of exclusion to about 95 percent. That number is the headline limitation: classical serology could often exclude a man, but it could almost never include one with the certainty a court wants. Modern paternity casework in India runs entirely on autosomal STR profiling; the classical methods now appear only in textbooks and historical case files.

Today every Indian DNA paternity test runs on autosomal short tandem repeat (STR) profiling read out on a capillary-electrophoresis platform. The science is mature, the per-locus mutation rate is well-characterised, and the math is standardised.

The Indian working panel. Most CDFD and CFSL labs use a globally validated multiplex such as the GlobalFiler kit (21 autosomal STR loci plus amelogenin for sex, DYS391 and Y-indel for Y-lineage screening). The 21 autosomal loci include the 13 original CODIS markers (CSF1PO, FGA, TH01, TPOX, vWA, D3S1358, D5S818, D7S820, D8S1179, D13S317, D16S539, D18S51, D21S11) plus newer high-discrimination loci (D1S1656, D2S441, D2S1338, D10S1248, D12S391, D19S433, D22S1045). The kit is run on a capillary electrophoresis instrument (the Applied Biosystems 3500 series is standard across Indian CFSLs and the CDFD), peaks are sized against an internal lane standard, and a software call is reviewed by an analyst.

The obligate-allele logic. For each locus you compare three genotypes: child (C), mother (M), alleged father (AF).

1. Identify the allele in C that did not come from M. That is the obligate paternal allele.
2. Check whether AF carries it. If not, the locus is an exclusion (subject to a mutation check).
3. If AF does carry it, compute the per-locus paternity index (PI), the ratio of the probability that AF is the father to the probability that a random unrelated man of the same population is the father.

The paternity index formula. For a simple case where the child is heterozygous and the obligate paternal allele has population frequency p, and AF is heterozygous carrying that allele, the per-locus PI is 1 / (2p). Multiply across all loci to get the combined paternity index (CPI). The probability of paternity W = CPI / (CPI + 1) at a prior of 0.5.

The Indian reporting thresholds.

• CPI greater than or equal to 10,000 and W greater than 99.99 percent: standard inclusion language. The alleged father is "not excluded" and the genetic evidence is "extremely strong support" for paternity.
• A single locus mismatch: scored as a possible STR mutation if the difference is one repeat unit and all other loci are consistent. Two or more mismatches across the panel are typically called as an exclusion.
• Mutation rate: the AABB and ISFG recommendation is to multiply the per-locus PI by the empirical mutation rate (roughly 0.001 to 0.005) at the mismatching locus rather than treat it as a hard exclusion.

For pedigree-level disputes (kinship, two-generation succession claims, identification of mass-disaster victims), the same STR data are fed into a likelihood-ratio framework that handles arbitrary pedigrees.

Most disputes are routine trio cases (mother, child, alleged father). NET likes to test the edge cases.

• Disputed maternity. Rare in India but it arises in baby-swap claims at hospitals, surrogacy litigation, and infant abduction recovery. The biology is symmetrical: the obligate maternal allele is identified and tested against the alleged mother. The same 21-STR autosomal panel and the same likelihood-ratio framework apply.
• mtDNA for the maternal line. Mitochondrial DNA is inherited only from the mother. Sequencing the HV1 and HV2 hypervariable regions of the control region (or, in modern Indian casework, full mitogenome sequencing on MPS platforms at CDFD and the CFSL Hyderabad DNA division) lets you test maternal-line relationships across many generations (grandmother, maternal aunt, sibling through the mother). mtDNA is not used for paternal-line questions because the father does not pass mtDNA to the child.
• Y-STR for the paternal line. The Y chromosome passes intact (apart from mutation) from father to son. A Y-STR multiplex (typically 17 to 27 loci including DYS19, DYS385, DYS389, DYS390, DYS391, DYS392, DYS393, DYS437, DYS438, DYS439, GATA H4) is used to confirm a paternal-line relationship when the alleged father is unavailable but a paternal uncle or paternal grandfather is. Y-STR cannot distinguish a man from his paternal-line male relatives, so it is a lineage tool, not a paternity tool, on its own.
• Motherless paternity test. If the mother is unavailable (deceased, refusing testing) the obligate paternal allele cannot be inferred from the mother's genotype. The math compensates by treating both child alleles as candidate paternal alleles and the threshold CPI is set higher to keep the false-inclusion rate low.
• Deficient paternity test. When the alleged father is unavailable (deceased, missing), Indian labs reconstruct his genotype from his confirmed biological parents (paternal grandparents) and / or his confirmed full siblings. The likelihood ratio is reformulated for the available pedigree; CDFD Hyderabad routinely runs these for property succession cases.
• Twin disputes. Monozygotic twins are genetically identical at every STR locus, so autosomal STR cannot resolve which of two MZ-twin alleged fathers is the biological father. Resolution requires deep whole-genome sequencing for rare somatic mutations, currently a research-grade workflow in India.

Three institutional layers are testable.

The referral lab. The Centre for DNA Fingerprinting and Diagnostics (CDFD), Hyderabad, set up under the Department of Biotechnology in 1995, is the national reference centre for DNA paternity and forensic identity work. It runs autosomal STR, Y-STR, X-STR, mtDNA and SNP-based ancestry panels, and produces the bulk of court-quality paternity reports in disputed civil matters.

The CFSL DNA divisions. The CFSL Hyderabad DNA division (and the smaller DNA units at CFSL Chandigarh, CFSL Kolkata and the newer DNA labs at CFSL Pune and CFSL Bhopal) handle criminal-side paternity work routed through state police, the CBI and the courts. The DFSS quality manual and ISO/IEC 17025 accreditation through NABL govern the workflow at every accredited Indian forensic lab, which is the framework documented in Indian forensic laboratories: CFSL, SFSL and RFSL.

State and regional DNA labs. Roughly a dozen state FSLs now have functional DNA units (Maharashtra, Tamil Nadu, Karnataka, Kerala, Gujarat, Uttar Pradesh, West Bengal and others), plus the regional FSLs in larger states. Capacity is the bottleneck rather than method; turnaround for a routine trio paternity case typically runs 4 to 12 weeks depending on backlog.

Research and reproductive-health work. The ICMR-National Institute for Research in Reproductive and Child Health (ICMR-NIRRH), Mumbai (formerly the National Institute for Research in Reproductive Health) and the AIIMS Delhi forensic medicine department publish much of the Indian peer-reviewed literature on population STR allele frequencies, mutation rates and pedigree-likelihood methods that the courtroom reports rely on.

The chain from sample to court is short. Buccal swab or EDTA blood is collected under judicial supervision, sealed, and routed with a forwarding letter from the magistrate to the lab; the lab follows standard chain-of-custody, runs the panel, computes CPI and W, and returns a signed expert report admissible under Section 39 BSA 2023.