Obligate Alleles

Cord blood, day two

Rahel Okonjo came down to the ward herself, which is how I knew before she said anything.

“There’s a problem with Nell’s panel,” she said. “It’s almost certainly ours. I’d like another draw off the two of you, and I’d rather do it now than have you hear a rumour at the water plant.”

Sennen types every child inside forty-eight hours. This is not suspicion, it is arithmetic: eleven thousand of us descend from six hundred and eleven embryos, and the mean kinship in the register has been climbing since the second generation the way a tide climbs, without malice, without pause. The register knows who everybody is a cousin to and at what remove, and it issues counsels — this pairing, not that one — and the counsels are why my grandmother’s generation has a word for stillbirth and mine does not.

Twenty loci. At each one you carry two alleles and your child must have taken one of them; that one is called obligate, because there is nowhere else it can have come from. Nell was missing mine at nine loci out of twenty.

“Nine,” I said. “Not twenty.”

“No. Nine.” Rahel had her hands folded. “If the sample were somebody else’s child entirely I’d expect worse than nine.”

Venous draw, day three

Same result, to the allele. Kes came out clean — Nell’s paternal half sat inside his genotype at all twenty loci, tidy as a docket.

Kes took this better than I expected and then worse. He sat on the end of the bed with his boots still on and said, “So it’s saying she’s mine and not yours.”

“It’s saying the sample isn’t mine.”

“Right, but Isla, they took it out of your arm.”

Buccal swab, day three

Also mine. Also excluded.

Rahel talked me through her list in the order she was obliged to work it, and she did not pretend the order was flattering. Sample mix-up: two births that week, both girls, and the register logs every barcode, and the barcodes were right. Transposition in the crèche: the cot tags were right. Undisclosed transfer of a stored embryo: my file said no, and my file had been kept by four different clerks over thirty-four years, so it was worth an hour of somebody’s time to be sure. Contamination. Mutation — nine loci at once, no.

“And then there’s the one I think it is,” she said, “which is that the panel is telling the truth and I’ve been reading it as one question when it’s two.”

“Meaning what.”

“Meaning the mismatched half of Nell isn’t random with respect to you. It shares about half your alleles. Whoever made that egg was your sister.”

I said I hadn’t got a sister and she said, “No. You haven’t.”

Hair follicles, four quadrants, day nine

They pulled six follicles from my scalp, six from my left forearm, six from the right, and a small punch of skin from above my hip. The forearms and the scalp were me — the me on file, the me in the water-plant roster and the crèche list and the counsels. The hip punch was somebody else. Fifty per cent shared alleles with the rest of my body. Sibling-level. The same stranger who had turned up in my daughter.

Rahel drew it for me on the back of a requisition. Two embryos were transferred into my mother in the spring of the fourth year — this was standard then, two, because the take rate was poor and the colony was young and impatient. One implantation was recorded. The other was marked not established, which is the register’s phrase for an embryo that went in and did not become anyone.

“They didn’t fail to implant,” she said. “They fused. Early — four days, five. Two zygotes, one body. It happens on Earth at some low rate and nobody catches it, because why would you look. Your blood came out of one of them. Your ovaries came out of the other.”

I remember asking whether I was ill. She said no, unusually firmly for her. “You’re two people’s cells getting on with it. They’ve got on with it for thirty-four years.”

Ovarian tissue, day twenty-two

I let them take it because I wanted the number.

The ovary was line B throughout — the other one, the one they had stopped calling not established. Not a mosaic in that tissue: all of it. Every egg I have ever shed came from a genome that has never been in my blood, was never typed at my own birth, and is not on the founder chart under my number.

So the true statement is this. Nell is my sister’s daughter. My sister has no blood, no scalp, no hands and no roster entry, and she has produced one child, and I carried it.

Kes did not like me saying it that way. He said, “You’re her mother, you were pregnant for nine months, don’t be so bloody exact.” Which is fair. Both statements are true and the register only has room for one of them.

Recomputation, week six

Here is the part I had not thought about, and Rahel had, on the first morning, in the ward, with her hands folded.

The counsels are computed from genotypes. Mine was wrong — not corrupt, not mislaid; faithfully recorded off my blood and simply not the genome that makes my children. So every kinship coefficient the register has ever computed between my offspring and anyone else in Sennen was computed from the wrong half of me. Every counsel I was given was an answer to a question about a woman who does not reproduce.

I was counselled away from someone once. I was twenty-three, he worked in the glasshouses, and the register put our prospective child’s inbreeding coefficient above the threshold on the strength of a shared great-grandmother. We were both sensible about it in the way you are at twenty-three, which is to say not at all, and then sensible in the way you are at twenty-five, which is to say entirely.

Rahel ran it again with line B. It came out under the threshold. Comfortably.

She told me because she had decided in advance that she would tell me anything I asked and would not decide on my behalf which facts were kind. I asked her not to tell me his number.

The recomputation went the other way too. Two of the pairings the register had blessed for my nieces — my registered brother’s daughters — were blessed on the assumption that they and my children shared a grandmother’s worth of ancestry. They share less. The advice was good advice for the wrong reason. Rahel is quite insistent about this distinction and I have come round to it. A system that is right by accident has not been tested.

Protocol amendment 41

The register now samples four tissues at birth: cord blood, buccal, heel skin, and a follicle pull. It costs eleven minutes and a little under two grams of reagent per child. Rahel put the amendment through in eight days, which I gather is a record, and she took the estimate to the council honestly: fusion at the observed rate, six hundred and eleven founders, double transfers standard for nineteen years — somewhere between nine and thirty of us are two people, and most of the ones who are will never have a reason to find out, because the second line is in a tissue nobody has cause to type.

They have not offered a retrospective survey and I have not asked for one. Nine to thirty people, told at fifty or sixty that the counsel that shaped their marriages was computed off a genome that isn’t theirs. Rahel says the arithmetic favours doing it. She also says the arithmetic is not the whole of it and that she is glad the decision is not hers, and both of those are the most emotional things I have heard her say.

Register entry, corrected

My number carries two lines now. Line A, typed from blood, 4th-year transfer, established. Line B, typed from gonadal tissue, 4th-year transfer, formerly not established, one offspring recorded.

Kes asked what I wanted to call her. Not Nell. The other one.

I said she already has something to be called. It’s on the founder chart, in the column of the ones that didn’t take, and it is four characters long, and I have known it by heart since the twenty-second day.