A Superior Court jury convicted the defendant of raping his twenty-one year old, mentally retarded and autistic sister. On his appeal from his conviction of rape, the defendant contends that the trial judge erred (1) in admitting expert evidence, derived from the results of deoxyribonucleic acid (DNA) paternity tests, that placed the defendant’s probability of paternity at 99.7 percent; and (2) in refusing to instruct the jury that probability of paternity evidence could not be considered evidence of intercourse. We affirm the conviction.
Background. The defendant was charged with rape after it was discovered the victim was pregnant and DNA tests performed on fetal tissue1 failed to exclude the defendant as a possible genetic father. The victim could not give valid consent to intercourse since, as stipulated by the parties and apparent from the evidence, she was “a severely mentally retarded woman who was and remains incompetent to make decisions on her own.” The Commonwealth’s evidence was that only two men had had the opportunity to impregnate her, the defendant and the victim’s father (also the father of the defendant). The DNA test results excluded the victim’s father.
During the summer of 1996, the victim lived at home with her mother, father, and the defendant, a high school student who would enter his junior year in the fall. She was “non-verbal,” with the capacity to use only a few simple words to communicate and the mental capabilities of a very young child. That summer, the victim received services from three female workers provided by Berkshire Family and Individual Resources (BFAIR). During the weekday, a BFAIR worker picked the victim up from her home or other prearranged location and would participate with her in activities within the community that included going to the mall, softball games at the park, the movies, or swimming. Due to the victim’s limitations, services *552were provided to her on a one-to-one basis and the workers were directed to keep her under constant supervision. The three women from BFAIR who provided services to the victim each testified that the victim was never out of sight of any worker and that she was never left alone with any other person.2 The BFAIR workers returned the victim to her family at 4:00 or 4:30 p.m. following the outings, most days bringing the victim to her mother’s workplace. A worker recalled only one day that she had brought the victim to the house and had delivered her to the defendant.
The victim’s mother, testifying on behalf of the defense, described the victim’s cognitive and other limitations resulting from mental retardation and autism. The victim weighed 176 pounds and was five feet, two inches tall; she was physically strong and could become combative if “somebody tried to get her to do something” (inferably, something she did not want to do).3 She trusted her family members, including the defendant, and displayed her affection by hugging them. The mother was largely responsible for the victim’s hygiene and dressed and undressed the victim daily, including changing the diaper-type underpants that the victim wore to bed at night. The victim’s mother observed no bruises or other marks on the victim during July and August, 1996. She testified that she and her husband, the defendant, and the three BFAIR workers were the only ones who had had direct contact with the victim during this period.
Both the defendant and his mother testified that, other than during a six-week period that summer when the mother at *553tended a class, the defendant was never alone with the victim. The mother testified that the victim’s bedroom was located upstairs near the parents’ master bedroom and that the defendant’s bedroom was downstairs, suggesting that the mother would have known if the defendant had entered the victim’s room at night. The defendant denied having sexual intercourse with his sister. He testified that he had rarely been home during the summer of 1996, because he had held two jobs and had worked sixty to eighty hours per week. From the beginning of July to the middle of August, 1996, however, the defendant stayed home alone with the victim for a short time (“ten minutes,” according to the mother), during two days of each week after the mother left for her class and before the victim’s father returned from work.
When a November, 1996, visit to the victim’s gynecologist revealed that the victim was pregnant, the State police were alerted. Due to the victim’s limitations, she was unable to communicate the circumstances attending her pregnancy, including the identity of the person who had intercourse with her.
A judge of the Probate and Family Court appointed guardians ad litem for the victim and the fetus. Pursuant to a court order that issued in that court following a hearing, the victim’s pregnancy was terminated in December, 1996. Dr. Laurent Delli-Bovi, who performed the abortion, testified that the victim’s hymen was not intact and that this was consistent with her having experienced sexual intercourse; that measurement of the fetal foot indicated the date of conception to be July 30, 1996; and that the date of the victim’s last menstrual period4 was consistent with a July 30 date of conception. Tissue samples from the fetus and blood samples from the victim were preserved, packaged and sent to Laboratory Corporation of America (LabCorp), a DNA testing laboratory in North Carolina. Tissue cell samples *554also were obtained from the defendant and from the victim’s father and sent to LabCorp.* ***5
According to Dr. Lloyd Osborne of LabCorp, an expert (in areas of immunology, human genetics and statistics) testifying on behalf of the Commonwealth, the DNA paternity test results excluded the victim’s father as a possible genetic father of the fetus,6 but did not exclude the defendant. Dr. Osborne further testified that, based on a statistical analysis of the test results, the defendant’s probability of paternity was 99.7 percent.7
Dr. Pravatchai Boonlayangoor, the defendant’s expert (in areas of medical microbiology, molecular biology, and immunogenetics), agreed that the DNA test results excluded the victim’s father as a contributor of DNA to the fetus. He adjusted the defendant’s probability of paternity statistic to 98.3 percent, on the basis of his assessment that because the victim, the defendant, and the victim’s father all shared the same genetic *555markers on two of the nine DNA systems tested, these two systems should have been discounted.8
Discussion. 1. Probability of paternity evidence. Dr. Osborne’s testimony as to the 99.7 percent probability of paternity statistic was admitted over the defendant’s objection. The only ground asserted at trial was that the evidence assumed a fact not in evidence, intercourse, and could not be admitted in the absence of independent evidence that the defendant had intercourse with the victim. He now argues, in addition, that the trial judge erred in admitting the probability of paternity evidence because such evidence infringes on the jury function to determine guilt beyond a reasonable doubt. The defendant does not argue that the statistical methodology employed in deriving the probability of paternity evidence was unreliable, nor could he do so.9 We first address the issue that was raised below.
“In weighing the probative value of evidence against any prejudicial effect it might have on a jury, we afford trial judges great latitude and discretion, and we uphold a judge’s decision in this area unless it is palpably wrong.” Commonwealth v. Siam, 434 Mass. 732, 752 (2001), cert. denied, 534 U.S. 1142 (2002). Cf. Commonwealth v. Lanigan, 413 Mass. 154, 166 n.13 (1992), S.C., 419 Mass. 15 (1994) (in a rape case, the materiality of paternity test results is “for the trial judge”). We *556do not think that the trial judge abused his discretion in permitting the testimony.
The evidence tending to establish that the defendant was the genetic father of the fetus was relevant to the determination whether the defendant had intercourse with the victim. See, e.g., Commonwealth v. Sicari, supra at 750-751 (semen found on back of car seat was relevant to issue at trial regarding extent of defendant’s participation in the actual killing of victim: “Semen is no different from fingerprints, hair follicles, or blood in its utility for this purpose. Such evidence connects him with the place, which in turn connects him to the crime that occurred there”). When a woman is pregnant, and there is no suggestion that she became pregnant through artificial insemination, that is evidence of her having had intercourse. The evidence of DNA testing of fetal tissue samples, like DNA tests performed on semen found in a victim’s vagina, is evidence of the defendant’s participation in the crime and is relevant.
We now consider whether the evidence, although probative, is unduly prejudicial. That concern was addressed in Commonwealth v. Beausoleil, 397 Mass. 206, 217 n.15 (1986), involving criminal proceedings to establish paternity.10 There the court concluded that inculpatory human leukocyte antigen (HLA) test results11 could be admitted to establish guilt in such proceedings and stated its preference that in so doing, “the HLA test results be presented to the jury in terms of the probability of paternity.” Id. at 217-218. Recognizing that the jury arguably would be “presented essentially with a statistical *557estimation of the defendant’s likelihood of guilt,” the court noted that probability of paternity evidence was not unlike evidence regarding the likelihood that fingerprint or voice print samples are from the same individual, evidence which also “is necessarily probabilistic.” Id. at 217 n.15.
In decisions since Beausoleil, our appellate courts have upheld the admission in evidence of DNA test results and related statistical analyses to establish a defendant’s presence at the scene of a crime. In Commonwealth v. McNickles, 434 Mass. 839, 847-848 (2001), the court held that it was not an abuse of discretion to admit testimony of the Commonwealth’s expert analyzing data using a statistical measure known as a “likelihood ratio,” which “compares the probability that the defendant was a contributor to the sample with the probability that he was not a contributor to the sample.” There, the vaginal swab obtained from the victim contained DNA from more than one person. Because testing does not always result in completely separating the DNA of the victim from that of the contributor of semen, the sample portion referred to as the “male fraction” may contain some of the victim’s DNA. The testing suggested, but did not confirm, the presence of the victim’s DNA in the mixed sample. The expert, testifying to the likelihood that the defendant was a contributor to the vaginal swab specimen, had to take into account the uncertainty whether it was the victim’s DNA that was, or was not, in the mixed sample. He testified that, assuming that the victim was a contributor to the mixed sample, it was 37,000 times more likely that the defendant was a contributor than that an unknown person was a contributor; assuming that the victim was not a contributor, it was 220 times more likely that the defendant was a contributor than that he was not. Id. at 847. The court rejected the defendant’s argument that the evidence should not have been admitted because of “the expert’s reliance on unsupported assumptions,” noting that the expert “presented statistical analyses for the full range of possibilities presented by the forensic evidence.” Id. at 848. The court concluded that the judge had not abused her discretion in admitting the analyses, “including an analysis that took into account the strong likelihood that [the victim’s] own DNA was in the sample taken from her vagina.” Id. at 849. Cf. Com*558monwealth v. Rosier, 425 Mass. 807, 813 (1997) (“Evidence of a [DNA] match based on correctly used testing systems is of little or no value without reliable evidence indicating the significance of the match, that is, ‘evidence of the probability of a random match of. . . the defendant’s DNA in the general population,’ ” citing Commonwealth v. Lanigan, 419 Mass. 15, 20 [1994]).
Here, “[t]he calculations on which the probability determinations were based were explained in detail.” Commonwealth v. Gomes, 403 Mass. 258, 275 (1988). During cross-examination of Dr. Osborne and direct examination of Dr. Boonlayangoor, defense counsel elicited the testimony that the probability of paternity estimate assumed a fifty percent likelihood that the defendant engaged in sexual intercourse with the victim; he highlighted the weaknesses of this assumption in his examination of Dr. Boonlayangoor and in closing argument.12 See Commonwealth v. Beausoleil, 397 Mass. at 220 n.19 (“We leave to defense counsel the task of highlighting ... the exact nature of the probability of paternity calculation and its weaknesses”); Commonwealth v. Girouard, 436 Mass. 657, 669 (2002) (no error in admitting expert’s opinion based on statistical frequencies that “no one other than [the defendant] is the donor of the DNA obtained from the sperm fraction of the vaginal swabs [of the victim]”; defendant’s trial counsel was “free to cross-examine the expert, or adduce rebuttal evidence, to establish that the DNA testing performed could not conclusively establish the donor of the sperm”). The prosecutor also placed no undue reliance on the probability of paternity statistic in closing argument. See Commonwealth v. Sicari, 434 Mass. at 752. There was no error.
We next consider the unpreserved claim that admission of the evidence relieved the Commonwealth of its burden of producing evidence of guilt and of persuading the jury of the guilt beyond a reasonable doubt. We need not decide whether the *559argument raises an issue of constitutional dimension, although it does not appear to do so. Our standard of review, under either view, is whether admission of the evidence created a substantial risk of a miscarriage of justice. Commonwealth v. Rivera, 425 Mass. 633, 637 (1997).
The additional evidence pointing to the defendant as the person who had had intercourse with the victim was compelling. The jury could have found that on or about July 30, 1996, the victim became pregnant as the result of sexual intercourse that occurred against her will; that the victim’s hymen was not intact, from which it could be inferred that she had engaged in intercourse; that only two men were identified as having had the opportunity during this time period to engage in sexual intercourse with the victim — the victim’s father and the defendant; and that DNA test results excluded the victim’s father as a contributor of DNA. This was sufficient other evidence to permit the jury to find beyond a reasonable doubt that the defendant had had sexual intercourse with the victim. See, e.g., Commonwealth v. Fowler, 425 Mass. 819, 828 n.20 (1997) (not prejudicial error to allow Commonwealth’s expert to testify that of the only four men who had contact with victim in the days preceding her death, DNA tests of semen in victim’s mouth eliminated all but defendant as source of semen); Commonwealth v. Fowler, 431 Mass. 30, 33-34 (2000) (DNA obtained from swab of murder victim’s mouth; presence of sperm in mouth of victim “was enough evidence to warrant the inference beyond a reasonable doubt that the victim had been penetrated orally”).
In addition, through direct and cross-examination of the experts and the judge’s instructions, the jury were informed that the probability of paternity statistic did not rely on evidence of intercourse, but assumed for purposes of the mathematical calculation that the defendant was equally likely to be the father as not to be the father.13
There was no erroneous admission of evidence. Even if there *560had been error, our analysis permits the conclusion that the risk was not substantial and that a miscarriage of justice did not occur.
2. Jury instruction. The trial judge also did not err in denying the defendant’s request to instruct the jury that the probability estimate could not be used as evidence that the defendant had had intercourse with the victim. The trial judge instructed the jury that they were free to reject or accept the opinions of the experts, and that they were to consider the soundness of the reasons underlying the experts’ opinions. They were specifically instructed: “if you choose to accept, in whole or in part, the test results and the probability of paternity, you are not to substitute any probability of paternity for the standard of proof beyond a reasonable doubt which I will soon explain to you. No mathematical probability can relieve you of your obligation to make that determination based on all of the evidence presented to you.” As the court observed in Commonwealth v. Rosier, 425 Mass. at 817, where the trial judge refused to give the jury an instruction with respect to DNA evidence that was requested by the defendant, “[tjhere is no requirement that a judge give a special instruction on the role of DNA evidence. The judge had discretion in determining how he would instruct the jury on the issue.”
In support of his claim, the defendant directs us to the court’s margin comment in Commonwealth v. Beausoleil, 397 Mass. at 220 n.18, where it is suggested that “the judge should instruct the jury, if ... so requested, that they may not consider HLA test results as evidence of intercourse, and that they may not consider such evidence of paternity unless they have found, beyond a reasonable doubt, that sexual intercourse at or about the time of conception had taken place between the mother and the alleged father.” In light of subsequent decisions of the Supreme Judicial Court, many of which we have discussed in this opinion, we do not think that this requirement would be *561extended to DNA evidence in the circumstances presented by this case.
In any event, as we have noted, the Commonwealth presented strong evidence, independent of the DNA evidence, that the defendant engaged in sexual intercourse with the victim. The Commonwealth, in its summation, did not focus on the probability of paternity evidence. The judge appropriately instructed that the jury evaluate the soundness of the technical principles underlying the expert testimony and that they were not to substitute the probability of paternity for the standard of proof beyond a reasonable doubt. Thus, any error on the part of the judge in declining to give the requested instruction was harmless.
Judgment affirmed.
The pregnancy was terminated.
One BFAIR worker recalled two instances at the BFAIR offices that she allowed the victim to remain in an office with the director and secretary while the worker used the bathroom across the hall. Another testified on cross-examination that, on occasion, she took the victim to the worker’s home that she shared with her boyfriend; there is no evidence that she ever left the victim alone with the boyfriend.
Indeed, Dr. Laurent Delli-Bovi, the victim’s gynecologist, testified that when the victim was brought to her office for an ultrasound to determine how far her pregnancy had progressed, the procedure took a very long time and was difficult because the victim did not understand what was happening to her. The mother, the doctor, and a nurse had to “work very hard” to get the victim to lie down on the examining table and keep her distracted so that the examination, which was not painful, could be conducted.
A BFAIR worker testified that the victim had her period on July 18, 1996; she recalled the incident because she had to locate a sanitary pad because the victim was not wearing one. The worker included the incident in her monthly progress report, a document that also was admitted in evidence. The mother testified that her daughter’s last period prior to her pregnancy was May 29, 1996.
The defendant and the victim’s father were ordered to submit to the taking of blood samples. See In the Matter of a Grand Jury Investigation, 427 Mass. 221, 226, cert. denied sub nom. A.R. v. Massachusetts, 525 U.S. 873 (1998) (“There is reason to believe [it may even be more likely than not] that either the father or the brother caused the pregnancy. Proper testing certainly will exclude one, and could exclude both, from the grand jury’s continuing interest. The test results, no matter what they are, will be a significant aid in the grand jury’s inquiry”).
A description of the scientific methodology underlying paternity tests that establish “the impossibility of the accused’s paternity ... to a medical certainty” is contained in Commonwealth v. Beausoleil, 397 Mass. 206, 209-210 (1986).
Dr. Osborne also described a second statistic relevant to determining paternity: the paternity index. The paternity index compares the likelihood that an alleged father could produce a child having the same genetic markers as the tested child with the likelihood that a random man could produce the same genetic markers as the tested child. The random man’s likelihood of paternity statistic is derived from gene frequency tables based on sample populations.
Dr. Osborne described the use of Bayes’ Theorem to calculate the probability of paternity, based on the paternity index data. He testified that the 99.7 percent probability statistic was based on what he termed to be a neutral prior probability of .5 — that is, that the defendant was fifty percent as likely to be the father as unlikely to be the father, not knowing anything further. He agreed, on cross-examination, that additional information could be incorporated to alter the probability statistic.
According to the evidence at trial, the probability of paternity statistic is applied to paternity test results which do not exclude the alleged father regardless whether the tests are based on human leukocyte antigen (HLA) or DNA.
Dr. Boonlayangoor testified that because “the mother, child, and the alleged father, all. . . have the same marker[s]” on the two systems (identified as the VWF and D5S18 systems), it could not be ascertained which marker was contributed by the father and which by the mother. He concluded that in these circumstances, the paternity index assessment with respect to each system (3.22 for the VWF, 1.79 for the D5S18) should be replaced with a factor of 1.0. Recalculating the paternity index based on these assumptions, the expert testified that the combined paternity index would be fifty-eight to one, or almost six times less than Dr. Osborne’s result of 334 to one.
The defendant did not request a voir dire to challenge the reliability of the process underlying the Commonwealth’s expert testimony. See Commonwealth v. Curnin, 409 Mass. 218, 222 n.7 (1991); Commonwealth v. Vega, 36 Mass. App. Ct. 635, 638-639 (1994). “To preserve objections to DNA analysis [on the basis of reliability], a defendant must file an appropriate pretrial motion stating the grounds for the objections and request a hearing in accordance with the principles set forth in Canavan’s Case, 432 Mass. 304, 309-312 (2000), and Commonwealth v. Lanigan, 419 Mass. 15, 24-27 (1994).” Commonwealth v. Sparks, 433 Mass. 654, 659 (2001).
The Commonwealth’s criminal complaint in Beausoleil was based on G. L. c. 273, § 12 (begetting an illegitimate child), now repealed. St. 1986, c. 310, § 25.
As described by Dr. Boonlayangoor, the HLA test uses the “ABO” blood group and the genetic makeup of white blood cell antigens to analyze forensic samples, whereas the DNA test looks at the genetic makeup of humans at the molecular level. HLA testing continues to be used in order to match tissue for transplantation. The primary drawback in HLA testing is that it requires fresh samples, whereas “the DNA structure in the cell is stable for a number of months and years” if collected and preserved properly.
See Commonwealth v. Beausoleil, 397 Mass. at 210 (“The HLA test is based on the identification and typing of more than fifty antigens found in the white blood cells. . . . When administered by itself, the HLA test excludes between 78% and 80% of all nonfathers; this figure increases to over 90% when it is administered in conjunction with the enhanced Landsteiner series”).
Dr. Boonlayangoor testified that the percent of probability of paternity “could be deceptive and misleading,” because it starts from a fifty percent prior probability and not from a one percent prior probability. “We presume immediately he is fifty percent, he is fifty percent [likely] to be the father. So it [is] deceptive from the beginning.”
We do not suggest that testimony regarding Bayes’ Theorem and the prior odds of .5 employed in this case would be admissible on facts other than those present here. See, e.g., Peterson, A Few Things You Should Know About Paternity Tests (But Were Afraid to Ask), 22 Santa Clara L. Rev. 667, *560685 (1982); Kaye, The Probability of an Ultimate Issue: The Strange Case of Paternity Testing, 75 Iowa L. Rev. 75, 93-94 (1989). Here, the assumptions employed by the Commonwealth’s expert to arrive at the probability of paternity statistic are consistent with the evidence in this case — that the defendant was one of only two men who might have been the father.