Argument

Orphan Genes Argument

Intro

If humans descended from a common ancestor we share with chimps, every gene in our DNA should have a chimp counterpart. The shared ancestor would have left the same genes in both descendant lines. That is the most direct prediction of common descent.

Real genomes do not behave that way. Humans have somewhere between 60 and 634 orphan genes, real working genes in our DNA that have no detectable match in chimps, gorillas, or any other primate. They are not broken leftovers (pseudogenes). They get transcribed into mRNA. They get translated into protein. Many are active in the brain. They have measurable phenotypic effects when knocked out. They are doing real biological work.

The phenomenon is not unique to humans. Every species sequenced so far has orphan genes. Fruit flies have orphan genes. Rice has orphan genes. Yeast has orphan genes. Estimates suggest 10 to 20 percent of any given genome is taxonomically restricted, present in one lineage and absent in all detectable relatives.

The mainstream response is de novo gene birth: orphan genes are produced when random non-coding DNA mutates into something the cellular machinery accidentally transcribes, translates, and selects for. The mechanism is real (papers like Carvunis et al. 2012 in Nature document it). The question is whether the rate of de novo gene birth scales to produce hundreds of functional, tissue-specific, biologically integrated genes in every species in available time. The rate-vs-time math is brutal. The probability of stumbling onto a functional protein sequence by random sampling of sequence-space is in the neighborhood of one in 10 to the 77th (Doug Axe, Journal of Molecular Biology 341, 2004). Producing dozens to hundreds of such genes from random non-coding DNA, integrated into existing cellular pathways, in 5 to 7 million years, is not a thing the math supports.

The argument here is not "we don't know how, therefore God." The argument is that the proposed naturalistic mechanism (de novo emergence from non-coding DNA) does not operate at the rate required to produce the observed phenomenon, and that independent design events (or front-loaded creation) explain the data more naturally. Common descent's smooth-gradation prediction has been falsified at scale; the orphan-gene pattern is what design predicts: discontinuity at the level of created kinds.

In full

The Orphan Genes Argument is an abductive argument from a structural mismatch between the smooth-gradation prediction of universal common descent and the observed pattern of taxonomically restricted genes (TRGs / orphan genes) in every sequenced species. Every species harbors a substantial fraction of orphan genes, protein-coding genes with no detectable homologs in other lineages, even close relatives. Humans have 60 to 634 orphan genes depending on detection stringency; orphan-gene fractions of 10 to 20 percent are typical across genomes. These are functional genes (transcribed, translated, expressed in tissue-specific patterns, with phenotypic effects). Common descent predicts smooth phylogenetic gradation; the orphan-gene data shows discontinuity. The mainstream rescue (de novo emergence from non-coding DNA) is mechanistically possible (Carvunis 2012) but rate-limited; producing functional protein-coding sequences from random non-coding regions at the observed frequency, given Axe-scale protein-sequence-space probabilities (~10^-77 per fold), is not what the math supports. Independent design events (or front-loaded creation unfolding at the appropriate times) explain the data more naturally. This page is structured as debate prep, each premise carries a second-order positive case, anticipated objections, rebuttals, a live-cite kit, and tactical notes. See Orphan Genes for the concept-side hub with detailed exegesis of the biblical and theological connections.

Argument structure

Form

Abductive (inference to the best explanation). P1 establishes the empirical phenomenon. P2 states the prediction of universal common descent. P3 examines the mainstream rescue and shows the rate problem. The argument is structurally identical to the Argument from Origin of Life form: the unguided-naturalistic mechanism (de novo gene birth here, prebiotic chemistry there) is rate-inadequate to produce the observed information-content. The conclusion is not that common descent is logically refuted; it is that the inference to design is at least as good as the inference to de novo gene birth at the observed rate, and on the cumulative case (combining with Edge of Evolution, Phylogenetic Incongruence, Cambrian) design wins. On standard IBE criteria, the design hypothesis fits the orphan-gene data better.

P1, Every species sequenced harbors a substantial fraction of orphan genes

Affirmative case (second-order arguments)

1. The phenomenon is universal across sequenced species. Tautz and Domazet-Lošo's landmark Nature Reviews Genetics (2011) paper documents orphan-gene fractions in fruit fly, mouse, zebrafish, Arabidopsis, yeast, and many others. Every fully-sequenced genome contains lineage-specific genes without detectable homologs in close relatives. The estimate of 10 to 20 percent of any given genome being taxonomically restricted has held up across additional sequencing.

2. In humans specifically, the orphan count is 60 to 634 depending on detection stringency. Toll-Riera et al. (Molecular Biology and Evolution 26, 2009) identified 271 putative human-specific orphan genes by BLAST. Knowles and McLysaght (Genome Research 19, 2009) confirmed 3 fully de novo human-specific genes with stringent empirical evidence. Wu et al. (PLOS Genetics 7, 2011) and Guerzoni and McLysaght (Genome Biology and Evolution 8, 2016) refined the estimates with different methodologies. Even at the most stringent threshold, dozens of human-specific functional genes exist.

3. Orphan genes are functional, not artifactual. They are transcribed into mRNA (verified by RNA-Seq), translated into protein (verified by proteomics), expressed in specific tissues (brain, testis, immune system), evolutionarily conserved among humans (under purifying selection, so doing biological work), and produce measurable phenotypic effects in knockdown and knockout experiments. McLysaght and Hurst (Nature Reviews Genetics 17, 2016) review the functional-evidence literature.

4. The detection threshold is permissive, not stringent. BLAST and related sequence-similarity tools are designed to find even faint similarity; orphan classification means homology fails even at permissive thresholds. The 60-to-634 range is conservative.

5. The pattern is more striking in lineages with rapid recent divergence. Brain-expressed orphan genes are enriched in primates; testis-expressed orphan genes are enriched across mammals; immune-system orphan genes appear in many vertebrate lineages. The pattern correlates with rapidly-evolving phenotypes.

Anticipated objections

1. "Orphan-gene counts depend on detection methodology; the headline numbers are inflated."
2. "Some orphan genes are mis-annotated as protein-coding; they may be non-functional sequences mistakenly identified."
3. "The phenomenon is interesting but explained within mainstream biology; you are inventing a problem."

Rebuttals

1. The lower-bound estimate is still substantial. At the most stringent threshold (only stringently-validated de novo human-specific genes counted), the number is in the dozens. Dozens of functional human-specific genes still require explanation; the high-end estimate is not load-bearing for the argument. Failure mode of the objection: chipping at the headline number while leaving the underlying phenomenon (dozens-of-functional-orphans) intact.

2. Mis-annotation is real but does not account for the bulk of orphan-gene findings. Knowles and McLysaght's confirmed 3 fully de novo human-specific genes (2009) used stringent multi-modal validation (transcription evidence, translation evidence, conservation across humans). Subsequent literature has applied similar standards. Mis-annotation accounts for some claimed orphans but not for the functionally-validated subset. Failure mode of the objection: using a real but limited concern (mis-annotation) to dismiss a broader well-attested phenomenon.

3. Mainstream biology treats orphan genes as an active research puzzle, not as solved. Tautz, Domazet-Lošo, McLysaght, Carvunis, and others are mainstream scientists working on the orphan-gene problem; the puzzle is real and acknowledged. The proposed solution (de novo gene birth) is itself the locus of the debate. Saying "mainstream has it covered" misreads the literature; mainstream has named the phenomenon and is working on the mechanism, not resolved it. Failure mode of the objection: conflating naming a phenomenon with explaining it.

Live-cite kit

• Scripture: Genesis 1:26-27 (humanity in God's image, categorically distinct); Genesis 1:20-25 (creatures after their kinds); Psalm 139:13-16 ("fearfully and wonderfully made").
• Scholarly: Diethard Tautz and Tomislav Domazet-Lošo ("The Evolutionary Origin of Orphan Genes", Nature Reviews Genetics 12, 2011); Toll-Riera et al. (Molecular Biology and Evolution 26, 2009); Knowles and McLysaght (Genome Research 19, 2009); Aoife McLysaght and Laurence Hurst (Nature Reviews Genetics 17, 2016); Carvunis et al. (Nature 487, 2012); Branko Kozulić (de novo gene calculations).
• Aphorism: "Every species has hundreds of genes belonging to no one else. Common descent did not predict this. The math has to account for it."

Tactical notes

• Lead with the universality. Every species has orphans. This is not a one-off anomaly; it is a general pattern.
• Cite specific papers. The orphan-gene literature is recent and technical; specific citation breaks the "ID is fringe" framing.
• Be ready for the mis-annotation move. Have the Knowles-McLysaght 2009 confirmation ready (3 stringently validated de novo human-specific genes).

P2, Common descent predicts smooth gradation

Affirmative case (second-order arguments)

1. Standard common-descent theory predicts continuity of gene families across phylogenetically related species. The reasoning: shared ancestors leave shared genetic content. If species A and species B descend from a common ancestor C, every gene in A should derive from a gene in C (or arise after the split); similarly for B. Both A and B should retain detectable homologs of the C genes. Failure modes (gene loss, rapid divergence) exist but are not predicted to dominate. Smooth gradation is the default expectation.

2. The phylogenetic-tree picture works because most genes do show smooth gradation. Across mammals, vertebrates, and even deeply into eukaryotes, most genes have detectable homologs in close relatives. The orphan-gene phenomenon is not the rule for the genome as a whole; it is a substantial minority pattern (10-20 percent) that contradicts the rule.

3. Francis Collins, Kenneth Miller, and the BioLogos defense of human-chimp common ancestry depends on the gradation prediction. Their case (shared genes, shared regulatory regions, shared chromosomal architecture) assumes the gradation rule. Removing the rule removes the foundation. The orphan-gene pattern is anomalous on the gradation prediction.

4. The functional-evidence-of-orphans intensifies the problem. If orphan genes were non-functional sequences, they could be dismissed as junk DNA that happens to differ between species. The fact that they are functional, integrated into species-specific biology, expressed in tissue-specific patterns, conserved within humans, means they are doing real biological work that has to be explained, not waved away.

5. The brevity of available time intensifies the problem for human-specific orphans. The chimp-human split is dated to 5-7 million years ago. Producing hundreds of fully functional, tissue-specific, integrated genes from scratch in that window requires either an unrealistic rate of de novo gene birth or an unrealistic rate of sequence divergence obliterating ancestral homology. Both push the math.

Anticipated objections

1. "Universal common descent doesn't predict strict gene-by-gene homology; it predicts overall phylogenetic signal."
2. "Gene loss is common; absence in chimps doesn't mean absence in the common ancestor.", gene-loss rescue.

Rebuttals

1. The "overall phylogenetic signal" framing is a goalpost-shift. Standard common-descent presentations (textbooks, BioLogos articles, the Theobald TalkOrigins case) explicitly emphasize gene-by-gene homology as evidence for descent. The "we never predicted strict homology" defense surfaces when the orphan-gene data is raised; it is not the framing used elsewhere. The argument identifies the goalpost-shift. Failure mode of the objection: shifting the prediction post hoc to accommodate the data.

2. Gene loss explains orphans-in-chimps-but-present-in-humans only if the gene was present in the common ancestor and lost in chimps. This would mean the genes we now call "human-specific orphans" should have detectable signatures somewhere in the chimp genome (pseudogenized, fragmented, or in non-coding regions). For most orphans, no such signatures are found. The gene-loss rescue would also have to explain why exactly the orphans relevant to human-specific traits (brain, immune, testis) were preferentially lost in chimps. The gene-loss explanation is mechanically possible for a small subset; it does not scale to the observed pattern. Failure mode of the objection: invoking a low-rate mechanism (gene loss leaving no trace) to explain a high-frequency phenomenon.

Live-cite kit

• Scripture: Genesis 1:24-25 (creatures "according to its kind"); Genesis 1:26-27 (humanity made in God's image, categorically distinct).
• Scholarly: Diethard Tautz and Tomislav Domazet-Lošo (Nature Reviews Genetics 12, 2011); Knowles and McLysaght (Genome Research 19, 2009); Francis Collins (The Language of God, Free Press, 2006); Theobald ("29+ Evidences for Macroevolution").
• Aphorism: "Common descent's prediction was: every gene has a relative. The data says: hundreds don't."

Tactical notes

• Use Collins's framing. Francis Collins's The Language of God (2006) explicitly uses gene-by-gene homology as evidence for descent. Citing his framing forces the opponent to defend it.
• Don't overstate. The argument is about substantial minorities (10-20 percent), not the majority of genes. Be precise.

P3, The de novo gene birth rescue is rate-inadequate

Affirmative case (second-order arguments)

1. De novo gene birth from non-coding DNA is the proposed naturalistic mechanism. Carvunis et al. (Nature 487, 2012) demonstrated that random non-coding sequences can occasionally become transcribed and translated, producing "proto-genes" of variable function. Some proto-genes do become integrated functional genes over time. The mechanism is real and documented.

2. The rate problem: functional protein sequences are vanishingly rare in sequence space. Doug Axe's Journal of Molecular Biology 341 (2004) experimentally estimated that approximately 1 in 10^77 random 150-amino-acid sequences fold into a functional enzyme. Branko Kozulić has developed the implications for de novo gene origination: producing a functional gene from random non-coding sequence requires sampling sequence-space at this probability density. The rate at which random non-coding DNA can produce a functional protein-coding gene is correspondingly low.

3. The time-budget cannot support the observed rate. For human-specific orphans (60 to 634 functional genes) to arise via de novo gene birth in 5-7 million years, the rate of de novo functional-gene origination would have to be approximately 10 to 100 per million years in the human lineage. Axe-scale probabilities make this unrealistic. The same calculation applies to other species: yeast orphan-gene counts, fruit-fly orphan-gene counts, all require similarly unrealistic de novo rates.

4. Behe's edge-of-evolution math constrains the rate. Behe (The Edge of Evolution, 2007) shows that random mutation plus selection can produce at most 2-3 coordinated mutations on realistic mammalian timescales, extrapolated from observed mutation rates in malaria, HIV, and E. coli. Producing a fully functional protein-coding gene from random non-coding sequence requires far more than 2-3 coordinated mutations (start codon, stop codon, ribosome-binding site, splice sites if applicable, protein-folding-capable sequence, integration into existing pathways). Behe's rate constraint applies. See Edge of Evolution Argument.

5. The proto-gene-to-functional-gene transition is the hardest step and the least documented. Carvunis's proto-genes are typically weakly transcribed, weakly translated, and weakly functional (if at all). The transition from proto-gene to fully integrated tissue-specific functional gene with a defined biological role is the contested step; it has not been documented at the rate the orphan-gene data requires.

Anticipated objections

1. "The de novo gene birth literature (Carvunis 2012, Tautz and Domazet-Lošo 2011, McLysaght 2016) treats this as well-established; you are dismissing mainstream science."
2. "Axe's protein-folding probability is contested by mainstream biology; the 10^-77 number is ID-movement framing."
3. "Homology detection has limits; many supposed orphans likely have homologs beyond BLAST detection.", the detection-limit deflection.
4. "This is just an argument from ignorance; you do not know how all these genes arose, but that does not mean design."

Rebuttals

1. The de novo gene birth literature establishes the mechanism, not the rate. Carvunis 2012, Tautz-Domazet-Lošo 2011, and McLysaght 2016 document that proto-genes can arise from random non-coding sequences and that some become functional over evolutionary time. They do not calculate the rate at which this mechanism produces fully integrated tissue-specific functional genes; they document its existence and discuss its frequency in qualitative terms. The argument concedes the mechanism and challenges the rate-vs-time scale-up. Failure mode of the objection: conflating existence-of-mechanism with sufficiency-of-mechanism.

2. Axe's 10^-77 number is peer-reviewed and replicated in concept. Axe's paper appeared in Journal of Molecular Biology (2004), a mainstream venue. Independent estimates (Reidhaar-Olson and Sauer 1990; Bowie and Sauer 1989) give similar orders of magnitude. The number is contested by some defenders of evolution (e.g., Hugh Hunt; Arthur Hunt of UKentucky) but the contestation is technical and does not establish a different order of magnitude. The argument's force does not require the exact value; it requires that functional folds are vanishingly rare in sequence space, which is well-established. Failure mode of the objection: labeling-as-ID-movement-framing rather than engaging the technical literature.

3. The detection-limit deflection works for a small subset of orphans, not for the bulk. For homology to be lost between humans and chimps in 5-7 million years would require evolutionary rates far above the observed neutral substitution rate of ~10^-9 substitutions per site per year. In 7 million years, neutral DNA shows about 7 percent divergence, far below the threshold at which BLAST loses signal. Even highly-conserved coding sequences retain detectable homology at greater than 70 percent identity after 7 million years. For homology to be obliterated would require catastrophic rates incompatible with standard molecular-clock data; such rates would themselves be evidence of non-Darwinian processes. Failure mode of the objection: invoking a limit of detection without demonstrating the orphan genes lie at that limit.

4. The argument is from rate-inadequacy, not from ignorance. "We do not know how" is not the claim. The claim is: "the proposed mechanism (de novo gene birth) does not operate at the rate required to produce the observed phenomenon (60 to 634 functional human-specific genes in 5-7 million years), given the empirical sequence-space probabilities (Axe 2004) and realistic mutation rates." This is a positive claim about mechanism-inadequacy, not a negative claim about ignorance. The structure is identical to the origin-of-life argument: the unguided mechanism's rate is insufficient. Failure mode of the objection: confusing rate-inadequacy with knowledge-gap.

Live-cite kit

• Scripture: Psalm 139:13-16 ("fearfully and wonderfully made... my frame was not hidden from You"; informational specificity of human formation); Genesis 1:26-27 (humanity in God's image); John 1:1-3 (the Logos as the universal source of life).
• Scholarly: Douglas Axe ("Estimating the Prevalence of Protein Sequences Adopting Functional Enzyme Folds", Journal of Molecular Biology 341, 2004; Undeniable, HarperOne, 2016); Michael Behe (The Edge of Evolution, Free Press, 2007; Darwin Devolves, HarperOne, 2019); Stephen Meyer (Darwin's Doubt, HarperOne, 2013, ch. 9); Branko Kozulić (peer-reviewed de novo gene calculations); Casey Luskin (Discovery Institute essays); Carvunis et al. (Nature 487, 2012); Tautz and Domazet-Lošo (Nature Reviews Genetics 12, 2011); McLysaght and Hurst (Nature Reviews Genetics 17, 2016).
• Aphorism: "The mechanism exists. The rate it would need to scale to does not. That is the orphan-gene problem in one sentence."

Tactical notes

• Lead with Axe's number when the opponent is mathematically literate. 10^-77 per functional fold is concrete, peer-reviewed, mainstream venue.
• Force the opponent to commit to a rate. Ask: "At what rate does de novo gene birth produce fully integrated tissue-specific functional genes? Is that rate sufficient for 60 to 634 in 7 million years?" The literature does not provide the answer because the answer the data requires is unrealistic.
• Don't bluff the proto-gene literature. Carvunis 2012 is real and documents the mechanism. Cite the papers; engage the actual debate. The argument's force is at the rate-scaling step, not at the mechanism-existence step.
• Pair with Edge of Evolution Argument. Behe's rate-constraint underwrites the rate-inadequacy claim. The two arguments together are stronger.

Conclusion

Orphan genes are best explained by independent design events, not by universal common descent under unguided naturalistic mechanism. Every sequenced species harbors a substantial fraction of orphan genes (10 to 20 percent of any given genome). Humans have 60 to 634 functional, tissue-specific, biologically integrated orphan genes with no detectable homologs in chimps or other primates. Common descent predicts smooth gradation; the orphan-gene pattern shows discontinuity. The mainstream rescue (de novo gene birth from non-coding DNA) is mechanistically possible (Carvunis 2012) but rate-limited: producing functional protein-coding sequences from random non-coding regions at the observed frequency, given Axe-scale sequence-space probabilities (~10^-77 per functional fold) and realistic mutation rates, is not what the math supports in the available time. Independent design events (or front-loaded creation unfolding at the appropriate times) explain the data more naturally. The orphan-gene phenomenon is one strand in a cumulative case against universal unguided common descent; combined with phylogenetic incongruence, the Cambrian explosion, and the edge of evolution, the cumulative inference favors design.

Master objections to the argument as a whole

1. "De novo gene birth is well-established mainstream science; you are misrepresenting the literature.", Reply: the mechanism is mainstream; the rate-scale-up to the observed phenomenon is what the argument disputes. The literature documents existence; it does not document sufficiency-of-rate. The argument engages the rate question directly.

2. "You're cherry-picking Axe's number; the protein-folding probability is contested.", Reply: Axe's Journal of Molecular Biology (2004) paper is peer-reviewed mainstream science; independent estimates give similar orders of magnitude. The argument does not depend on the exact value but on functional folds being vanishingly rare in sequence space, which is well-established.

3. "This is just God-of-the-gaps.", Reply: it is rate-inadequacy-of-the-mechanism, not knowledge-gap. The structure parallels the origin-of-life argument: the unguided mechanism is insufficient, design is at least as good an explanation. The conclusion is positive (design as best explanation given evidence), not negative (we don't know, therefore God).

4. "Even granting orphan genes, common descent is supported by other independent lines.", Reply: conceded; the orphan-gene argument operates within a cumulative case. It targets one specific anomaly. Combined with Phylogenetic Incongruence Argument, Cambrian Explosion Argument, Edge of Evolution Argument, the cumulative inference favors design.

5. "You're conflating absence of detectable homology with absence of homology; some homology is below BLAST sensitivity.", Reply: addressed in P3 rebuttal 3. For homology to be lost in 5 to 7 million years would require catastrophic rates incompatible with the observed molecular clock. The detection-limit deflection works for a small subset, not the bulk.

6. "This doesn't prove young-earth or separate-creation; it just shows current mechanisms are inadequate.", Reply: conceded. The argument concludes to independent design events (consistent with multiple Christian readings), not to a specific creation chronology. The narrowing to a specific Christian model comes from convergence with other evidence.

Tactical opening / closing

Opening line: "If humans and chimps share an ancestor 7 million years ago, every human gene should have a chimp counterpart. The shared ancestor would have left the same genes in both lines. Reality is messier. Humans have somewhere between 60 and 634 functional, working, brain-active, immune-relevant genes that have no detectable counterpart in chimps, gorillas, or any other primate. They are not junk. They make proteins. Some are essential for development. Where did they come from in 7 million years on a random-mutation account? That is the orphan-gene problem, and it is not a fringe concern; it is documented in Nature, Nature Reviews Genetics, Genome Research, and PLOS Genetics."

Closing landing strip: "The Orphan Genes Argument does not say evolution is impossible. It says the unguided-evolution-plus-selection mechanism (de novo gene birth from non-coding DNA) does not scale to the observed phenomenon. Every sequenced species has 10 to 20 percent lineage-specific functional genes. The math (Axe-scale probabilities for functional folds, Behe-scale rate constraints on coordinated mutations) does not support the production rate the data requires. Independent design events, or front-loaded creation unfolding at the right times, explain the data more naturally. Common descent predicted smooth gradation. The genome says discontinuity. The data fits design as a positive prediction, not as a gap to be filled."

Connection to Scripture

• Genesis 1:26-27, humanity in God's image; categorical distinctiveness corresponds to substantial unique genomic content (brain-expressed orphans, cognition-relevant orphans).
• Genesis 1:24-25, creatures "according to its kind"; orphan-gene patterns mark the boundaries of biological kinds.
• Psalm 139:13-16, "in Your book they all were written, the days fashioned for me"; informational specificity of human formation parallels the genomic finding of substantial human-specific functional content.
• 1 Corinthians 15:39, "all flesh is not the same flesh"; categorical distinction between human, animal, fish, bird; molecular-level distinction documented by orphan genes.
• John 1:1-3, "all things came into being through Him"; the Logos as source of biological information.
• Colossians 1:16-17, "in Him all things were created"; Christ as designer of the unique biological content of each created kind.

See Orphan Genes (concept hub) for fuller exegesis of biblical and theological connections.

Patristic / scholarly note

Classical / patristic / medieval:

• Basil of Caesarea (Hexaemeron, c. 378), each kind reproduces "after its kind"; categorical distinction between kinds.
• Augustine (De Genesi ad Litteram, c. 415), rationes seminales; the kinds were given their distinctive seed-content at creation, unfolding in time.
• Thomas Aquinas (Summa Theologica, I, Q. 73), distinct species in God's creative work; categorical biological distinction.

Modern, mainstream:

• Diethard Tautz and Tomislav Domazet-Lošo, "The Evolutionary Origin of Orphan Genes", Nature Reviews Genetics 12 (2011): 692-702, the landmark review.
• Aoife McLysaght (Trinity College Dublin), multiple papers defending the de novo gene paradigm.
• Anne-Ruxandra Carvunis et al., "Proto-genes and de novo gene birth", Nature 487 (2012): 370-374, the mechanism paper.
• Sara Knowles and Aoife McLysaght, "Recent de novo origin of human protein-coding genes", Genome Research 19 (2009): 1752-1759.
• Yuan Wu et al., "De novo origin of human protein-coding genes", PLOS Genetics 7 (2011): e1002379.
• Adam Guerzoni and Aoife McLysaght, "De Novo Genes Arise at a Slow but Steady Rate", Genome Biology and Evolution 8 (2016): 1222-1232.
• Aoife McLysaght and Laurence Hurst, "Open questions in the study of de novo genes", Nature Reviews Genetics 17 (2016): 567-578.
• Eugene Koonin (The Logic of Chance, Princeton, 2011), proposes de novo emergence at the broader OOL level.

Modern, ID-side:

• Stephen Meyer (Darwin's Doubt, HarperOne, 2013, ch. 9), frames the orphan-gene problem within the broader information-genesis critique.
• Michael Behe (Darwin Devolves, HarperOne, 2019), engages the orphan-gene literature from an ID perspective.
• Douglas Axe ("Estimating the Prevalence of Protein Sequences Adopting Functional Enzyme Folds", Journal of Molecular Biology 341, 2004; Undeniable, HarperOne, 2016), the protein-folding probability work underwriting the rate critique.
• Branko Kozulić, de novo gene-origin probability calculations.
• Paul Nelson and Ann Gauger, ID-side analysis.
• Casey Luskin, "How Orphan Genes Pose a Problem for Common-Descent Theory" (Discovery Institute, multiple articles), ID-side commentary on the literature.

See also

• Orphan Genes, concept-side hub with detailed biblical / theological exegesis and broader concept-level framing.
• Common Descent Critique, concept-side parent.
• Edge of Evolution Argument, rate-constraint twin argument; positive-evidence-twin to this negative-evidence argument.
• Phylogenetic Incongruence Argument, parallel argument from incongruent gene trees.
• Common Design vs Common Descent Argument, underdetermination argument that the orphan-gene data feeds into.
• Cambrian Explosion Argument, parallel rate-vs-time argument at the phylum level.
• Protein Sequence Space Argument, Axe's protein-folding probability underwriting the rate critique.
• Specified Complexity Argument, the information-theoretic design filter.
• Argument from Origin of Life, parallel rate-inadequacy argument at the abiogenesis level.
• Adam and Eve Historicity, closely related to the human-distinctiveness question.
• Intelligent Design, framework.
• Stephen Meyer, primary contemporary defender.
• Origins and Science, concept-side master.
• Origins, master hub.
• Arguments, top-level master index.